Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Walter Brattain by Charles Weiner on 1974 May 28,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Early experiences in science at Whitman College, Washington, from 1920; friendships with fellow students and teachers. Graduate study at University of Oregon and Harvard University; difficulties funding education; study with Edward A. Milne at Oregon and John Van Vleck at Harvard. Work at National Bureau of Standards on piezoelectricity and oscillators; work at Bell Labs on thermionic emission and experimental basis of statistical mechanics; influence of Arnold Sommerfeld on his work on the copper oxide rectifier. World War II work with National Defense Research Council on the magnetic head of submarine detectors. Return to Bell Labs following World War II; research in solid state with group headed by William Shockley and Stanley O. Morgan; preliminary researches in semiconductor effects.
I’m Walter Brattain, and it’s Tuesday, May 28, about 9:30 in the morning, 1974.
This is Charles Weiner, and we’re sitting in Professor Brattain’s office at Whitman College —
Walla Walla, Washington.
These I must return to you now.
OK, thank you.
More about them later. Well, let’s point out that this is really a continuation and a supplement to an interview that was conducted with I guess Alan Holden and Jim King many years ago, which we each had a chance to glance at recently. But there are questions that have occurred to me since then. Also some of them came to me from looking in your personal archives. I’d like to start in the thirties. I was looking at your publication list, and I find that if this is an accurate list of your bibliography, you have publication No. 4 in 1934, and the next listed is No. 5 in 1941. That’s an important gap. I wondered what kind of period this represented for you in your work?
What does 1941 say there?
It’s copper oxide varistor in the Bell Labs RECORD. Now, I was curious about the nature of the work in those intervening years. Was this a kind of a long slow process which did get results, but nothing that you felt was appropriate for publishing in PHYSICAL REVIEW? Or you may not have been able to publish it because it had to do with company policy? I don’t know if it was either of those.
The story here has two parts. At the time, we were essentially working in competition with Walter Schottky of Siemens-Schuckert, and often Becker and I found that, about the time we had something in regard to copper oxide that we thought might be worth publishing, we would receive an article from Germany that Walter Schottky had already published. This is mentioned in the other story, a little bit. On the other hand, Alan Holden, who was then in the publications department, came out and interviewed us, and wrote up, after interaction back and forth with us, something about the copper oxide work that would go in the record. And we thought this was pretty good. But Oliver Buckley, who was then possibly some place in the organization up the line — he later became president of the Bell Telephone Laboratories — had a feeling that one should protect the Bell Labs from publishing too soon, and he clamped down on this article of Holden’s at the time. We did not particularly like this attitude, because we thought that this was against what we felt was the general feeling of the Bell Labs about publishing scientific work. Of course, this was not a scientific story, it was on the level of what is published in the RECORD. This went back and forth, and finally in the later years — I don’t know when, the date is there in the RECORD, isn’t it?
It was published in ‘41, the copper oxide.
Yes. Finally we were allowed to publish in THE RECORD. We got through to people (a publication in THE RECORD that does have a lot of the scientific work) what we were doing on the copper oxide. It was only after the war that I published an article in the REVIEWS OF MODERN PHYSICS, 23, 203, 1951. It was a memorial issue, in honor of J. T. Tate. I was asked if I wanted to contribute to this. I published quite a story on the work that we’d done during this period. The difficulty, however, in this period was that, copper oxide being such a messy type of structure sensitive thing. Theoretical ideas were that the semi-conductor conductivity was due to a missing copper ion in the lattice –- the impossibility of determining a missing ion, lack of perfect stoichiometry in the lattice — when you have exactly the right number of both copper and oxygen in the lattice was not possible to determine. At the level of a few parts in ten million, let’s say, or parts in a hundred million, to determine how many missing copper ions there were was almost impossible. So what we did was finally published.
When did Holden first raise this question of writing something up?
It must have been about five years before the publication in THE RECORD.
Around ‘36. In other words, there was enough done by that time. Well, that would be where the progress reports for the group might be helpful. I could take a look on that. For example, about ‘34 —
When is the first time copper oxide is mentioned?
It’s mentioned in the year ‘32-‘33. I don’t know if it’s mentioned before.
No, it wouldn’t be. It couldn’t be. It’s got to be after that.
In this one, the progress report for August 3, 1931, you say, “Some five weeks was spent at University of Michigan attending lectures given by Professor Sommerfeld on Fermi-Dirac statistics and their application, problems of conduction and thermionic emission from metals. It was planned to continue this study until a thorough knowledge of this method of attack is obtained.”
This was the purpose. We were then studying the thermodynamics and the statistical mechanics of the work function in thermionic emission, and the photoelectric work function, and we published a theoretical paper finally. [J. D. Becker & Brattain Phys. Rev. 45, 694 (1934) note reference in this paper to Schottky’s paper Phys. Zeits. 20 49 (1919)]
Yes, that was the paper of ‘34 which —
— yes —
An interesting thing I came across in your files was that Becker first wrote to Williams at Michigan about arranging for you to come and attend a lecture, although as an employee of the Bell Labs you couldn’t get the entire summer off, but you could come to the first part of the summer. You wanted to hear Sommerfeld’s lectures, and you preferred to share accommodations with a group of advanced graduate students. The next thing attached to that in the file is a memo, an internal memo to Becker’s group, saying that you have attended those lectures and that you will be giving a course of 16 to 20 lectures of your own on Fermi-Dirac statistics, based on what you learned in the Sommerfeld course. I find that very interesting, that there was a real connection with what was going on in the academic world, and then translating it into internal Bell Labs lectures.
I have my notes of the lectures I gave. This was the first opportunity at Michigan for me to go and listen, without worrying about details, but looking at the overall picture from Sommerfeld’s lectures, and I didn’t take notes — oh, minor notes. But when I got back to the Laboratories, the Laboratories said, “Well, now, you’ve learned all this, how about telling us about it?” And I had to scramble to get the material from which I could really get the details. It was very interesting that people like Quarles (who later went down to Washington) were in this group. The first lecture I gave, I found I wasn’t well prepared, and I struggled through the whole lecture. The next lecture, I had down pat. I just went bang, bang, bang through it, and after the second lecture, any number of people in this group said that my first lecture was the best one, because they were all trying to help me. So I learned right there that if you were a good enough actor in giving a lecture, you should act like you were having trouble, because this gets participation out of a group. But of course you should have it all well in hand.
This was Becker’s group itself?
No, no, anybody in the Labs who was interested could attend this.
It was on Thursday afternoons for about one hour and it started sometime in the latter part of October, and I guess it was once a week and 16 to 20 sessions meant it went practically through the equivalent of an academic year. What was the average attendance?
Oh, I would say, on estimate, at least at the start, thinking about the size of the room at least 20. Some of them were executives. Quarles was an executive. They were interested. I started in with the classical statistics then Fermi-Dirac and Bose-Einstein. I started in with probability. I can show you these. There were nine lectures.
We’ll look at them later.
These people were all generally physicists, weren’t they?
Well, and engineers.
But they kept up?
Oh yes, but there was no exam.
This was on Bell Labs time. What about your preparation time?
That was Bell Labs time too. I worked like a slave. It’s an interesting comment, that my first thought was that I could get the information I need out of Darrow’s ADVANCES IN COMTEMPORARY PHYSICS. But when I started reading Darrow I found out that while he did a very nice job of giving you a qualitative feeling for it, the details weren’t there.
This was still when Bell Labs was in New York?
In New York, yes.
Bell Labs people could have, and did, participate in colloquia at various universities, NYU or Columbia?
But there was nothing quite like this going on at those places, is that right?
No. It was Sommerfeld who first applied Fermi-Dirac statistics to conduction in metals. Classical statistics could not explain this. His work was only a few years old at this time!
Did any people from local universities come in?
No, no, this was strictly Bell Labs. I’m quite sure Buckley was in this group — I don’t know whether Kelly had time or not.
Well, we started off on the progress report. I guess we were sidetracked. Without careful study, the first one that seems to mention copper oxide, at least begins with it, “Case 19-881, Copper Oxide,” and that’s in the progress report for December 1932-33. And so the reason we consulted this is to try to figure out how long the work was going on which was eventually covered in this 1941 article.
Well, let me express what the problem was, as far as Becker and I were concerned, at this time. From the literature there was a confusion as to where the seat of rectification was, whether it was at the interface or phase boundary between the copper oxide and the copper, or whether it was in the body of the oxide. Selenium rectified in the other direction. The selenium rectifier and copper oxide were the two commercial rectifiers at that time. And there was argument as to whether the conductivity in the body of the semi-conductor was ohmic or not. One of the first things that Becker and I did was to prove to ourselves scientifically that the internal conductivity in the semi-conductor was ohmic, and therefore the rectification had to be at the interface, and the next question was: which interface? In both semi-conductors. And it was between the oxide and the copper on which the oxide had been grown, in the case of copper oxide. In the case of selenium, the rectification was after the selenium layer was formed on some kind of metal. Selenium melts at about 110 degrees. In the case of selenium it was between the selenium and the contact you put on the selenium afterwards. And those things we tied down. Those were the first things we tied down. There was great confusion in the literature about this.
The reports are signed by Becker generally, some by Becker, some by you. But specifically, the thing I’m looking for — let me pin it down. I just had it here a minute ago. OK, in the report for the year 1935, “Outline of Work in Department 328, Section G-2, J. A. Becker, W. H. Brattain.” I don’t understand, Department 328 was what?
It was our department.
What was its English language name?
Well, it was in the research group — what was the date again?
This is 1935.
This was probably under W. Wilson. He was director of research. Arnold died in 1933 and Wilson replaced him.
So Department 328 would be just a research group. Would it be a research group on semi-conductors?
No, not necessarily. It was I believe Wilson’s department.
The section, Section 6-2 would be what? What you call a group?
I don’t know. That was probably our section.
It says, “Work done since last conference,” so the implication here is that there were annual conferences to report the progress of the year.
Yes, well, there were conferences when we thought we had something, or somebody wanted to know. At this time, the people in the metallurgical group under E. E. Schumacher were producing the copper oxide rectifiers. I can’t think now of the name of the person who was actually doing the work on copper oxide. It was G. O. Smith. There were a lot of things involved. Chile copper was the only copper that would make good copper oxide rectifiers, mined in Chile, I think in the oxide zone. We were also doing another thing in this period — it’s mentioned in here. You had to have another contact, and it used to be aquadag, plus a pressure contact with a lead washer. And one of the things we did was to find that under proper conditions, with proper sandblasting — lightly, so as not to create damage due to fracture in the oxide layer — a light sandblast and then a silver or gold contact to the oxide would be ohmic.
You’ve got two phase boundaries in a semi-conductor, and the least understood, even till after the war, was how you made an ohmic contact to a semi-conductor, a contact that would not rectify at all. And I don’t know whether those memos which were on file have been destroyed in the Bell Labs, or are still in existence. We found that we could make an ohmic contact to copper oxide, such that you couldn’t detect any rectification at the contact, by evaporating silver or gold on a freshly sandblasted surface of copper oxide. And we went into the technological development of the manufacture of rectifiers with such contacts. We made a great big steel vacuum chamber in which you could put the trays of copper oxide rectifiers around the inside surface of the chamber, and a tube of tantalum wrapped with silver or gold in the center. And by using liquid air traps, actually in the chamber, you could get a pressure on the order of 10 to the minus 6 mm of mercury in a matter of ten minutes after you closed the chamber. You had to get down to something of the order of less than 10 to the minus 5 to get this ohmic contact by evaporation. We took this instrument over to Western Electric, and in fact, I got a little carbon tetrachloride poisoning over there, because we used to use carbon tet to clean. I was in a closed room, and I had to get down on the floor to wash this damn cylinder, and carbon tet goes down, it’s heavy, and I got a stomach ache. But by a large amount of work, it finally went into production at Western Electric. That was in that same period. (1937) About 10,000 copper oxide varistors were processed at the laboratories before moving it to the Western Electric plant at Kearney. J. A. Decker from Kearny assisted in this work.
It’s a very rich, busy period.
I found that in the vacuum system — well, you see it up here in this picture, up there in the top corner. That was the bell jar system I had. The problem was, how do you get low pressures in a system that you’re going to have to open and close? I did a lot of research on that, and the result of the research was that if you have a good diffusion pump, which we bought from Henderson at the University of Washington, an oil diffusion pump, if you have that diffusion pump running, the limiting pressure is the pressure at which there’s an equilibrium between vapors coming off the walls and the pump speed. And the pump speed that you can get from a pump, even if the opening to the pump is about four inches in diameter, is the order of something like — I can’t remember now — 40 liters per second probably or something like that. But these things that came off the walls, that made this equilibrium, were things that would condense upon a liquid air trap, and if you put a two inch bulb inside, which you could force liquid air into, right inside the vacuum system, this little two inch bulb, for anything that was condensable on it, would have about 1000 liters a second speed. And you could pump down to a certain level. You’d have a liquid air trap in the pumping line, and get down to that level, and you could force liquid air in to the bulb and come down, as I remember, almost immediately to a pressure ten times or more lower. It’s just an equilibrium problem. This was never published, but people questioned me about this. And this has been used in technology.
Beyond Western Electric you mean?
Yes. What we did was put a liquid air trap in the reverse end of this big steel tank, a metal liquid air trap. We didn’t want any glass in the system, and it had a big fin. The liquid air trap was just a steel tube. It was a question of conductivity and heat capacity, because you used up a lot of liquid air to cool it, and then a great big fin covered the whole end of that cylinder, so the liquid air trap was actually in the cylinder. Well, these are things that we just found out, investigating what caused the limiting pressure.
Was that work included in this memorandum that’s referred to in the outline of work, Department 328, the year 1935, under the category, “Work done since last conference,” a memorandum entitled, “The results of a study of the current voltage characteristics of copper oxide rectifier units have been completed.”
That wasn’t this.
No, I don’t mean this, but is that the thing you referred to before, the write-up that was not released? Or was this something —
— well, it may have been about this time. (1937)
This was ‘35 — it says it’s completed, doesn’t say what happened to it. Then you continue on about the copper oxide, copper blanks in vacuum, before oxidizing them, pre-glowing them before oxidizing them. You continued to work with the vaporization apparatus. First, you find a suitable metal or combination of metals to use in making a good low resistance ohmic contact —
— OK, OK.
Second, you use the apparatus to investigate making synthetic rectifiers, semi-conductors, or electron conducting materials of all types that might be of some future use. This work was done in my bell jar vacuum system in which I did the first work on silver and gold contacts to copper oxide varistors.
Well, R. W. Sears at this time was working in the same group under Becker, and he was working on silicon carbide. And there again, we looked for conductivity being ohmic in the body of the silicon carbide. But in silicon carbide there were phase boundaries, like PN junctions, that were hard to investigate. We did not know about PN junctions then.
Which you didn’t fully understand at the time.
Here, in the outline of work in the same department, work accomplished in 1937, you say, “An apparatus was designed and constructed to vaporize metal contacts onto copper oxidize varistors on a commercial scale.”
Yes, that’s it.
“The apparatus has been working for the past six months, and is soon to be transferred to Kearney” — was that Western Electric?
Yes, Western Electric.
“It was designed to take 3/16, 1/2 inch, 3/14 inch, and 3 by 5 inch copper oxide varistors. That comes down to the time of the story you’re talking about. Right. These documents are important. So we go on — this is worthwhile — there are memos referred to. And then, March ‘38, you indicated that the work on the vaporization apparatus was continued. You changed to a single tube of molybdenum or tantanum as an evaporation means, and so forth.
Well, that helps to fill in this period.
It’s interesting to note that when this equipment was used at Kearney, the material that we used to sandblast the surface of the oxide accumulated in the vacuum chamber, and I once saw it, after it had been used for a few years, and it just flabbergasted me, that they could still do this, with the amount of this damn dirt in the vacuum system.
And it didn’t affect it really?
No. They finally made themselves a vacuum system — oh, it must have been almost after the war, I guess, in which these trays went through the vacuum system, through chambers continuously in which there were a sequence of pumps along the system.
Well, it figures it would be the next step ultimately. That’s really on a production scale.
Yes. I had nothing to do with that. In fact I was flabbergasted that they could do it.
The report for 1940 is very interesting. This was submitted by you on May 28, 1940, and you say, at the very beginning, “In the past year, a qualitative picture of the nature of the contact between a metal and a semi-conductor or two semi-conductors has been developed,” then you start to talk about, “a potential hill occurring in the region of contact” and so forth. It seems that this is of a different order than the things you —
— is silicon mentioned there? You see, this is after the old —
I’ll take a look at it… “experimental work on the varistors, report of conference in Fletcher’s office.” You might talk to me about Fletcher too, that’s another man I want to talk about.
Well, this is after Mott, the first Mott and Schottky articles. It developed, as I remember, and I can’t say which was first, that we began to realize that there was a direct analogy between the rectification at the surface of the semi-conductor, and the rectification in a vacuum tube diode. Mott’s paper predicted that the rectification would be dependent on the difference in work function, the contact potential between the semi-conductor and the metal contact. And that in a sense, it was emission of electrons or holes from the semi-conductor into the metal that was responsible for the rectification. The reverse flow from the metal was small and not too dependent on the applied potential. This is after Wilson’s model of the semi-conductor.
A good deal after.
How did you learn of that work? Did it come to you through the literature?
The literature. Oh yes. Mott’s paper was the first one we saw.
So the extension of that is what’s reported on —
Well, give me something, without going into all the details, of the flavor of —
At this time Debye was a consultant at the Bell Labs, and there was some discussion on these things with Debye, and this is the time that Becker had his idea that didn’t work.
We were talking yesterday about this, that Becker’s style was to be rather stubborn about an idea, unless you really could argue him out of it.
Yes. Well, more than that. I guess later in this period we did have separate offices, but for some time we had an office together, and if I would make a suggestion to Becker, or even write out a little thing, Becker did not understand it until he worked it out for himself. And he had a tendency that sometimes annoyed me to think, after he’d worked it out for himself, that it was his own idea. But I’ve said certain places, and I want to say this again in this context, Becker dried my ears off as a green young Ph.D. Becker could open an article, PHYSICAL REVIEW, when it first came in, when we’d get it in the office, find an article that we thought had something to do with any area of our work, and he could kind of thumb through the pages without reading in detail, and put his finger on the weak spot, just like that. He was tremendously critical, on anything. If I’d say in one of my reports that “this is obvious,” he’d say, “How obvious?”
That must have been very helpful to you.
Oh yes. He made me very critical. Taught me lots of things.
What was the difference in age between the two of you?
Oh, about three or four years. Yes.
While we’re on the subject of Becker, yesterday you found a document — a letter from M. B. Long which sort of recalled your experience when he first interviewed you. Why don’t we take a look at that, because that relates to the earlier interview and clarifies some of the issues of how you came to Bell Labs. You thought it was — well, you tell the story.
All right. Well, in the first place, I think I first met M. B. Long at Whitman College, on one of his trips. The second time I met him was in Oregon. He had in tow at that time Jacques Cortell who still publishes AMERICAN MEN OF SCIENCE, and he wanted to be sure that Jacques Cortell got a chance to play tennis while he was at the University of Oregon, and I played tennis with him. I later met his brother socially in New York when I was there. Then, the next point is when I applied in the spring of ‘28, for either a job at Bell Labs or at the Bureau of Standards. Apparently Long was in on this, from his letter, but he didn’t act soon enough. This explains Tate’s telling me, when I got the telegram from the Bureau of Standards, that I should wait a little while, I’d get an offer from Bell Labs, which I didn’t know about then. Tate knew about this, see, but he didn’t mention Long’s name or anything to me. Long and Tate were good friends.
Then the next contact with Long was when he came down to the Bureau of Standards, and by what I thought was accident, I was the only man to show him around the radio section. And I told him about the work we were doing and showed him the standards, capacity and inductance, gave him an overview, mainly in my area. This may not have been accident. He may have asked to see me. He told me then, any time I wanted a job at the Bell Labs to let him know. But I thought he was looking, had in mind my working in this area at Bell Labs. And I did not want to be a radio engineer. It was at the meeting of the American Physical Society in April 1929 that I walked over to the electrical measurements building, met Tate, and he introduced me to other members of the Physical Society who happened to be there, and the last person he introduced me to was Becker. And you see, Long’s letter explains that he had in mind originally my working with Becker.
And that he was waiting, because of the internal protocol to go through channels, wasn’t that right?
Well, no. You read his letter. It doesn’t imply quite that, does it?
I looked through it very hastily. I looked through, why he was so apologetic, that’s the important thing. This is a letter written in 1972 from Long to you.
Yes. “In the late twenties, at a Physical Society meeting at Bureau of Standards, L. O. Grondall presented a short paper on copper oxide rectifiers. He included no theory, but I was impressed with the potentials, and I was pleased to accept a couple of samples. Then I discovered that patent rights had been offered to the Western but had been turned down on advice from the newly created Bell Labs. This made my blood boil, because I found that Arnold had been in on the turn-down, and I talked to Jack Ling telling him of some of the things I’d already done with the samples, finally commenting that some characteristics were so similar to those of vacuum tubes that we ought to find out what made them tick. As I had hoped, he was sufficiently impressed so that he talked with Arnold, winding up with, ‘Arnold, are you going to investigate this or am I going to have to.’ Arnold agreed, and later told me that Joe Becker would carry on, and gave me authority to hire a physicist with Ph.D. training.
So when I first met you, I had been and was looking for someone to fit the situation. After talking with you at length and at length with Jack Tate, whom I had known intimately… “see, he saw me at Minnesota, I’d forgotten this —” I came to the conclusion that you were ideally fitted by temperament as well as by training, and knowing Joe Becker, I was very anxious to have Joe make the decision, and right there and then was where I made my great big unintentional, amateurish blunder. Engineers were given verbal offers, to be confirmed later, but I told you nothing. In the meantime, I convinced Joe, but before I had time to write an offer, I received a letter from Jack Tate that made me suddenly realize how unfair I’d been to you.” See, this is after I accepted the offer from the Bureau. “I have regretted that incident, but there didn’t seem more that I could do at the time. We can at least be thankful that it had no other effect in your feelings toward me. But I want you to know that your accomplishments have given all of your associates a great deal of vicarious satisfaction.”
All that was written by Long.
Yes. See, I didn’t know this.
Did you have negative feelings toward him about it?
No, no, no, no! It probably was better for me that I went the way I did. Oh no. I wrote him back a letter, and there’s another letter here that he replied to that.
We’re talking about 23, 24 years later from the event. Well, that helps to fit in the origins of some of the work. These things maybe they’d be important to copy, as a supplement to the interview. They relate directly to it. We’ll talk about that later. What I mean by that is the progress report and then perhaps that letter. The issues that I’d like to get on to now have to do pretty much with the question of patents in general, about the problems that someone like yourself would face in an industrial laboratory, doing basic science, with results that are basically important to physics, wanting to maintain professional standing in the field... I was starting to ask about the problem of a professional physicist working in an industrial environment, about relationships to the larger professional community and responsibility to the company, especially as it comes up in issues of public knowledge of what you’re doing. And it comes up it seems to me in another issue which flows from that, the question of patents. So let me first talk about the patents. I’m not clear what patent rights individuals in Bell Laboratories have, how that works out.
You are paid a dollar when you become a member of the Bell Telephone Laboratories for your patent rights.
For all of your patent rights. Any time in the future. I see. But does that mean, you sign an agreement at that time?
Well, you’re paid.
In advance, but you still apply for the patent in your name?
Western Electric applies for the patent in your name.
It’s clear in the application that you have assigned your rights to them.
Then who’s the one who decides on whether something is patentable?
Of course, the Bell Labs grew like Topsy out of an organization that was formed in Western Electric during World War I, when there wasn’t any vacuum tube, and the furthest you could talk by telephone was by shouting over loading coils, from about New York to Denver. And communications were important. This is really a story in itself. I’ve given a talk on what made the Bell Labs tick, and I have the notes of it here in my files. I looked up all the people that were important to me, or I became acquainted with in the Bell Labs, when I first went there in ‘29, which is only four years after the Bell Labs was organized under Jewett. And practically all these people went to work for Western Electric during the war years. J. B. Johnson of Johnson-Noise, Davisson, Germer — I have a whole list of them here. I can’t mention off hand all of them. Then after it was organized various people, scientists who were in this group, went into various jobs. The head of publications for a long time was Mills, and I think he was a trained scientist. Certain things became customs.
One of the interesting customs that I think had a large influence at this time was that once you got a request for apparatus OK’d up the line and purchased, it was your apparatus. Nobody could take it away from you. You could be asked if you’d loan it, if you were through with it, when somebody else wanted to use it. You took better care of the apparatus that was yours than would be taken in a central bureau. It also made it possible for people in the Labs who had ideas that the management was not enthusiastic about, to borrow apparatus from their friends and go ahead and work on it. And in this lecture that I gave to a Rotary Club down in Portland, I said that in my estimation, the Bell Labs was proudest of those things that were done in spite of management. The thing is in Long’s letter, a little bit. All right. Now, let’s go to the other part, the patent question. In the thermionic emission papers that we published, as I remember there was no particular trouble, though I was under Becker and he may have carried that end of it. But it developed that whenever you published a paper, the Patent Department was notified, and was given a chance to write a patent covering this.
In the convention countries, you had a certain length of time to apply for a patent after the paper was published. There was no difficulty. In very important patents, though, you worried about a non-convention country, such I think as Argentine, where somebody could listen to your paper and try to beat you on an application. But this only came up in patents that looked like they were tremendously important. Usually we didn’t worry about the non-convention countries. So, on the other hand, the ability to publish in certain areas was influenced by the personal convictions of people who happened to have gone up out of science into management. And Buckley was a little bit cautious in this area. For one reason or another. And that’s what happened, partly, in the copper oxide area. But the other thing was that Schottky beat us, for years. In fact, this is one of the first things I said to Schottky when I met him, “I finally caught up with you.”
Well, were there many instances other than the copper oxide, in your own experience, where something that the scientists would consider worthy of publication in the open literature would either be delayed, held up for some reason, or prevented from publication?
I think there was a period under Buckley when there was a tendency to make it a little bit difficult.
A lot of gripes, morale problems developing because of this?
I wasn’t aware of any particular ones. Becker and I felt entirely in a position of fighting this tooth and nail. But we had to have a good case. Our reaction was, we’ll fight, not that we’ll quit. There’s another thing about the Bell Labs organization. Well, maybe this is not the place for it, but you very quickly learned that you had the freedom to call anybody in the Bell Labs, at any level of the organization, without going up through channels and back down the other way. If you came across something that you thought was important for the Bell Labs to know, and you knew the individual it was appropriate to tell this to, you called him and told him directly. And it was experience in the Bell Labs that I never called anybody (by telephone, at any level of the organization, even the president of the Bell Labs, and found he was not there or busy on the phone or in a conference, and left word with his secretary that I’d called) that that individual did not call me back. Except once, and that was Honaman in the publications department. He had an assistant call me back. I never called him again. Found somebody else in the publications department to call. In fact, that pretty near made Fisk mad once. We were discussing something in my office. This was after the transistor. I forget what it was. And he said, “I think you ought to call Honaman.” I said, “Fisk, if you want to call Honaman, go do it.” He did. This is another tradition in the Bell Labs. In the Bureau of Standards, everybody was known by initials. You initialed reports when things came around. WHB, I was known as. At the Bell Labs, everybody was known by their last names.
How would you address Buckley, for example?
I’d say “Buckley.” Oh, I might say “Dr. Buckley” at the higher level, but — There was one time, after the Bell Labs management read this book on HOW TO WIN FRIENDS AND INFLUENCE PEOPLE, that the management started calling everybody by their first names. There was a long period of time when Fisk was away from the Labs. In fact, we gave Fisk his party when he left the Labs to go to Harvard. I’ve tried to find those details lately and I can’t find them any place. I was right in the middle of this. And I think the first time I saw Jim Fisk he was walking one way across the passage between two buildings and I was walking the other, and he comes up smiling and says, “Hello, Walter, how are you?” And I said, “Fisk, when you call me that, smile.” It was annoying to me, because I learned everybody in the Bell Labs by their last names, and I have a poor memory, and to learn the first names over again when I was older was just completely impossible.
But it was meant to cut down barriers, I guess.
Oh yes. Sure. But I thought the old tradition of just calling everybody by their last name, Becker, Johnson, so on and so forth, no titles, except during World War II. Whenever there were armed forces people in the place, we made it a practice to call everybody Doctor. You can see the reason why. Well, you know, we felt they needed to be impressed.
A question related to the one about professional publications, you say that people weren’t too annoyed, you and Becker at any rate fought on it. What was your expectation in terms of career? You went there in 1929, went there with good recommendations from good physicists and with some sense of your own worth. Did you expect to stay there the rest of your career?
Yes. I wanted to be a research physicist.
You could have done that at a university too.
There was a period under Fletcher — when Fletcher, back in the late 1930’s became director of research — in which I was told by Fletcher that there were opportunities out in the Labs where they had jobs for men. It meant a rise in the organization chart and it meant of course an increase in salary. And I told him very definitely that I wanted to be a research physicist. I of course wanted to get paid what I was worth — I probably told this very softly to him, because Fletcher was a very nice individual. Fletcher reported this to Kelly, who was next in line in the organization at the time. Kelly didn’t accept this and called me up into his office and we had a long discussion. I told him, Kelly talked strongly and the way to get along with Kelly was to talk back to him. I told him very strongly that if I couldn’t be a research physicist in the Bell Telephone Laboratories, I’d have to find another job. And, of course, if I’d accepted one of these opportunities for advancement, I wouldn’t be sitting here now.
You mean, in what way?
Well, let’s take one of the men who came in after the war, K. G. McKay that did the fundamental work, along with another colleague, on the use of semi-conductors to detect high energy particles, which is now used throughout nuclear physics, you know. He was offered a job, an executive job, after being my supervisor for a while, best supervisor I ever had. He knew better, he could say better what I was trying to do than I could. He was offered another position in the development area, took it, he went right up the line just like that, became vice president of AT&T, is now executive vice president of Bell Telephone Laboratories. The Laboratory, by malice aforethought, has in essence maintained a group of about 500 from the time I went in — it developed a little bit, it wasn’t quite as clear-cut then — of people whose primary job was to understand physical or scientific phenomena, of any interest to the communications industry, including now work on genetics, which is information theory. But they always hired into that group more people than needed to maintain the 500, because scientists, chemists, physicists who were inoculated with the idea that the only career was scientific research, would not come to the Labs unless hired into this group. But after they were in the Bell Labs and saw the opportunities, the Bell Labs continuously made competent engineers and executives out of some of this group.
And their policy was to do it from within.
Presumably I was the second individual to refuse to go up the organization line. Davisson was the first. He didn’t talk like I did, and just out and out refuse it, but he was Becker’s boss when I went to work for Becker in the organization. Davisson soon developed a method of not taking those things out of his In-Box that he wasn’t interested in. Did I ever tell the story about Davisson’s press conference when he got the Nobel Prize? Is that in here?
How he turned to you and said, “Someday you’ll get it.” Right. During this period — first of all, when was this conversation with Fletcher, then Kelly?
Well, can you find out when Fletcher was…?
I mentioned it in terms of something you said — let’s take a look — progress reports will show it, they always have…
…this is before World War II.
Here’s one, 1941, “Report of Conference in Mr. Fletcher’s office.”
It was probably a few years before this.
This one says, 1940 it’s in his office, then previous one is Becker, so — l940.
Yes. It’s about that time.
During this period of the thirties, did you have second thoughts about the possibility of academic opportunity to do research?
Assuming you were in touch with universities, you were asked to give talks at various colloquia and so forth…
…that may be in my letter files, correspondence files, offers. But I never had an offer that I was interested in. I had many offers, especially after the transistor, but even before. But they were mainly offers to guide things, not to do research myself. Fletcher, of course, was one of the executives that carried on, while he was executive, his own research in acoustics. But this is very hard to do. Once you went up about two steps above where I was, your job was to take care of the people that were working under you, if you were going to do the job right. In fact, I made it a policy that even my technical aide reported to my boss and not to me. And I never had a professional man work with me that reported to me. He always reported to my boss, because I’m a very direct person, and I wanted to have him feel that he was free to talk back to me. And of course, this was one of the questions Becker asked when he hired me, as to whether I’d talk back to him. He was tired of having underlings that would listen to him and then go on and do it their own way, without arguing it out. Becker and I would sometimes argue a situation, bang, bang, bang, and then after we’d gone for a little we’d convince each other, and we’d start arguing the opposite ways.
Switch positions. What about Fletcher, how was he to work with? Did you have the same kind of relationship with him?
A man of that kind, and of course he was doing his own research, he was two steps up the line, and he was busy. Not that he was not easy to contact, but when you want him, it’s hard. And he had an executive assistant, which was W. S. Gorton who wrote the story (he was by training a physicist), and I soon found that an easy way to get things done, or get messages up the line, was to call Gorton. He was available, he’d make a note of then, at the appropriate time he’d bring them to Fletcher’s attention. Fletcher was a very kind individual. In fact, Kelly had a tendency to lack respect for any man in the organization who didn’t do things the way he did them and he thought Fletcher was too easy going. He never got over this. I considered this somewhat lacking in Kelly’s character, because I came to the conclusion pretty early that there are as many ways of doing things as there are individuals, and just because the individual doesn’t do it your way doesn’t mean that he’s no good. But when they’re working with me, they’ve really got to convince me. In this context, I never sat down and built an apparatus with the research in mind, a complete apparatus. Homer Hagstrom did, successfully. My way of doing it was to set up the simplest experiment and make a trial run. The trial run showed me that I wasn’t getting what I wanted and I’d improve the setup. And we’d go along. There was one instance, in working with Garrett, in which after he came with me I think we worked for a year and a half, filled the notebooks with the data and always we found that we weren’t getting the information we wanted. And after a year and a half we finally had the proper setup so we were pretty sure we were measuring what we wanted to measure. We had the setup we wanted. We spent the whole day looking at a ‘scope — qualitatively, order of magnitude, checking signs which were important, being sure that we knew the signs of the effects and when they changed. Garrett went away and wrote the paper, and I spent three or four days afterwards getting the actual experimental magnitudes.
But you knew…
We knew what we were doing. I didn’t have to change any words in his paper. I just had to put in the exact numbers.
I want to get to your working relationship with him in a little bit, in a different time sequence. Let me take you to a completely different realm. That is your relationship with neighboring universities in the New York area. Here you were in the laboratory, probably a larger concentration of physicists in that one laboratory than at any specific university. I don’t know but it may have been. What kind of network of contact was there with people at Columbia, NYU and various places in the city of New York?
There was a fellow-graduate, L. P. Granath, from Minnesota who was teaching at New York University at the Heights, and he asked me once up to give a seminar, up there. This was generally O.K.ed by word of mouth — there was very little opposition. We were encouraged, we could go and talk about our work. Usually they would be about things that had already been in memos or were about to be published or something like that. We had notices of Columbia University physics seminars that came to us directly, and we went to many. While we were in New York we attended many of those seminars. If they looked interesting to us we went up, on company time. In fact, at first there were time clocks for a certain level of organization, when we went into Bell Labs. But I wasn’t on it. Research scientists were off the time clock, and this eventually disappeared entirely.
In the thirties?
It was more likely that the people from the Lab would go to the university, rather than the university people coming down to the Lab?
Well, we had a colloquium that was organized before I came in. K. K. Darrow could tell you more about this, you should ask him. We had people from all over. Anything of interest. This support financially was by the Bell Labs, but the Bell Labs management really had nothing to say about the people we invited in. It was an organization of a group of scientists in the Lab, and these were fine colloquia. If you could get any of the minutes or anything out of Darrow of that organization, you ought to get it in your files — I’ve been after him on this. For one period I was elected secretary. A few of us used to take our guest speakers over to Charles French restaurant on Sixth Avenue for dinner after his talk. Sometimes outsiders were invited to hear these talks on an informal basis.
He’s going to turn over his papers to us, as far as I know. But he has no idea what’s in them.
He’s never had the energy to dig in and find them. If they’re still preserved, or any records of this, they will be in his papers.
That would show who spoke, on what subjects. You mentioned one seminar at Columbia, we talked about yesterday. I’d like to get the details of that. So I could check the missing…
Well, Fisk and Shockley would know more details about this. It’s my memory that we attended a colloquium or seminar at Columbia University, and Bohr at this seminar read a telegram — I think the telegram was from Meitner. Let’s see, there was Hahn and Meitner and there was…
Frisch would have been. See, Hahn and Strassman were still in Germany, Frisch and Meitner were in Scandinavia.
Yes. It was Meitner. It was a telegram from Meitner that fission had been obtained. And I know, because we’ve talked about it among ourselves, in the Laboratory, I mean Fisk and Shockley, these were members of the group — and they immediately started work on calculating how much of uranium they’d have to put in a pile, to have the reaction go. I can remember distinctly their announcing, they thought they knew the size of the pile. I can even remember just interacting with them, maybe a group seminar — as to where it would be safe to do this. Whether it would really be safe to do it anywhere on the earth. The whole earth might go. The realization of the impact of this was immediate, and the concern about the possibility that Germany might use this as a military weapon, and be able to blackmail the world. It was immediately evident, to any physicist, and the physicists agreed not to publish their results. They put on the classification before government ever did. But they did send around their papers, did spread the information of what they were doing around in the group who were working on this. And I know, just hearsay, of the effort to get to President Roosevelt, and it was finally necessary to get a name scientist Einstein to do it. He would not listen to anybody else!
Getting back to the Columbia colloquium, was this something that you had gone up for on the basis of one of those notices?
And it was at that colloquium that this announcement was made. Then the calculations that Shockley and Fisk did were back at the Bell Labs?
They were back at the Bell Labs, and a patent was issued, was filed by Western Electric on their work, for a pile. And I know for a fact that all these applications went in the secrecy file, and then when they were opened up after the war, the earliest date was Shockley and Fisk. And Western Electric finally decided to withdraw the patent, because, for one thing, I don’t think Shockley and Fisk had any idea of the role of impurities in the absorbing of neutrons and so on and so forth. And they thought it was only proper that the people who actually did the work should — well, it was a log jam, and there was going to be trouble if Western didn’t withdraw.
Had there been any earlier discussion of the possibility of Bell Labs doing work in nuclear physics, early in the thirties?
I can’t remember the times, but probably only after the war. There was from time to time discussion as to whether we should have a cyclotron or some facility, and it was always decided that it was really something that the Bell Lab shouldn’t try to do. I don’t know what facility they have now. This was fought back and forth, but it was a rather big project, and it was something that they decided was not too worthwhile. They have availability, I think now, to obtain trace material on everything they need in their work.
I came across a letter from Shockley to Merle Tuve in the late thirties, ‘37, ‘38, regarding the possibility of Bell Labs getting into nuclear physics by getting some kind of accelerator, and debating the various merits. That’s the only evidence I have, and someday I should ask Shockley — both, mostly Shockley —
Tuve is pretty sick, isn’t he?
Well, he’s getting around somewhat.
Tuve is one of my favorite characters. I met him at the Bureau of Standards first. We had a joint seminar there, at the Bureau and the Department of Terrestrial Magnetism. I remember that’s the first time I ever head Henry Norris Russell, the astronomer. He was a wonderful person, too. In fact, that’s the only thing, in the days before I became famous, that I ever regretted — that I hadn’t instead become an astrophysicist, because they get to travel. I kind of wished that I’d become an astrophysicist.
For that reason.
Yes. Well, travel, and I was always interested in astrophysics.
It’s a little after 11, can we keep it up?
I think we can. Maybe you’d better, if you can stop the tape, I’d better mention — two people to go to lunch.
I think we can bring this part to a close before then. The thing that I want to get into now is another question about working in groups at Bell Labs. Once a group was established, how strict was it in terms of its composition? Was it possible to co-opt other people? What kind of communication was there among groups?
Oh, there was plenty of cross-communication between groups. I might just describe the organization, well, I have described the organization of the group after the war, haven’t I? I have also described, that the fact that we had difficulty getting the chemists to release a physical chemist to our group. But we brought pressure and got him. This was Gibney, who was of course an essential part of our group. But on the other hand, Schaff and Thererer — the metallurgists — furnished us with our semi-conductor material, and I talked with them, anytime we wanted to. But the close working together was in the group. If there came a problem, and we thought we knew somebody in the Labs who was an expert in this, we could ask for help.
What if someone wasn’t asked but had some feelings about what you were doing? Could he really penetrate, come in with his ideas? Was it easy for a person to do that?
We’d listen to them. The management would — well, for example, Kelly immediately called Becker and I on the work Ohl was doing at Holmdel on silicon. And from then on, we cooperated with Holmdel. I started on silicon. This was before the war. We’d go down to Holmdel and Holmdel furnished us a setup so that we could put our point contact detectors right into their wave guide, when we went down there. Furnished us with all the things we needed, point contacts and everything else. All we had to do was mount our little sample and put it in there.
This is OK at the stage where it’s just a regular group working on a project, but when you’re onto something that you think may be major, cause a revolution in the industry itself, which is very proprietary, it will affect your ability to interact with another group. Let me ask this, when was the first inkling that that would be the case in the surface state project that you were working on?
Being the transistor?
We published the surface state work.
Yes. You did some of that with Shockley. Did you recognize, when you published those papers in 1947, that this was very significant?
It gave us a completely new concept, a new theoretical structure, to design our experiments around. We immediately started testing, trying to measure the space charge layer that we knew was there in the semi-conductor, in equilibrium with the surface. We didn’t quite get that simple idea of the statement at that time, that developed really. At least I began to realize this, slowly. Let’s go to the other period after this, immediately after the discovery and the demonstration of amplification by injection of minority carriers by another point contact, immediately this was company confidential, it was even more than that. There was a group that was informed of this, including H.S. Black, who immediately went to work, getting material from us, and making transistors, point contact transistors, and studying them. So there was a large group inside the laboratory that was informed. Also, and this was purely my feeling — I had a definite feeling, from previous examples in which it looked like there was a breakthrough, being advertised up the management line very quickly, there was a tendency not to report this at higher levels until they’d waited a few weeks. It was so damned important that they were scared, if they told Kelly about it, that it might be a flop. And they didn’t want to advertise it until they were convinced themselves that it wasn’t a flop.
“They” meaning people at the research lab level.
Yes, it would maybe be Ralph Bown and it would be the people who saw the demonstration. The names of those are down, in Gorton’s story. But it very rapidly was dispersed throughout the development area. Well, Becker was in a separate group, and they were actually in a position to be working on something close. Becker and Shive I think were pretty close to making a germanium wedge with which they could put point contacts on the two sides of the germanium, which, if they’d done it before our discovery, they would have probably discovered this themselves. And they interacted with us immediately.
The timing — the memo that I saw yesterday, did I copy that or not? I have some notes on it. Typed memo, I guess it was signed by Shockley, and the title was, “Concerning the Report on Semi-Conductors,” and it gives the schedule. It was dated December 17, 1947. It gives the schedule of reports, to be presented in two parts, December 17 and December 23. The one in which you were involved, for example, which was Part B of that, was the first one, December 17. How much was known up until that time, when that memo was given by Shockley, was that already, the realization that you were there, practically?
The first observation of hole injection was on this sample, in which we evaporated a gold film, by going around with the point contact on the side of the film, and finding out an effect which was the reverse — as far as sign was concerned — of the field effect which we hoped to get. That was the first, and that was either December 16th or 17th. The demonstration, the reduction of practice as you know, was on the 23rd, 24th.
We were doing a chronology, you remember.
Well, it is my suspicion, I don’t know, that that memo you speak of was probably restricted in distribution. It probably didn’t go too high in the organization.
Well, while you’re talking let me get it. I think I know where it is. Here — unless we pulled it out yesterday?
No. I was looking. As a result of the 25th anniversary, we went through a lot of these, trying to find them.
I did pull one out yesterday, I remember now as a matter of fact, I pulled it out and you took it — here it is right here. There were two memos. The first is the one I’m talking about at the moment. Now, there’s the distribution right there.
Yes. This is pre-transistor. Everything that’s mentioned in here.
That’s what interests me. It’s dated December 17.
Yes. Well, it takes time to make a memo. That is a report — “A present tentative schedule for the report on semi-conductors is as follows: a proposed program to be given on Tuesday afternoon, December 23rd.” Well, that is referring to the demonstation.
Then this other one that they refer to on December 17th, during Professor Debye’s visit, you see, he was there at that time.
Yes. This was probably not totally disclosed by that time. We, I believe, discussed with Debye the results that Gibney and I had obtained.
The people who got this, if you look at the list, does that imply anything about limited distribution, or would that have been the normal distribution?
Well, no management names are on here, but Bown has been told about this. He was presumably directly above Shockley, as director of research or possibly vice president in charge of research.
So those names are pretty much the group.
“Surface states” became the code name, under the classification.
So the Part B, do you recall that meeting, what you presented on the 17th? It was Part B which included your experiments relating to surface states, was that scheduled for December 17th?
This is definitely Gibney and my explanation of the field effects, using electrolytes. It’s the best I can tell you on that.
Well, the next memo below that is on April 21st and it’s regarding a conference on the present status of the surface state job — the code name for the transistor.
Probably — well, now, look here...
There are two of them. I was referring to this one, first, which is separate from the other.
Now, Gibney has left. Sparks has joined the group as a chemist to replace Gibney. But this is the group. It’s being called by Bown to report.
Again signed by Shockley, as head of the…
Now, the next memo below that is on May 6th, and it regards the policy of confidentiality and just how reports should be handled and so forth. But it seems pretty late to me, confidentiality, on May 6th.
Well, it was by word of mouth, before this.
Now it’s gotten to the point, apparently — in the first paragraph it says there’s a need for an increasing number of reports, and therefore there are these procedures, as to who should get them. So this was in reference to not only external confidentiality, which I assume you were observing all along, but to codify the procedures of copies of documents.
Even way back, when Gibney and I first got the field effects, then Bardeen and I produced an amplifier using the field effects, at very low frequencies, and I told my writing group about it that night, and then after that, I called up Bardeen and I said, “We should tell Shockley what we did today,” and we called him that night. But the next night, when I went over to my riding group, I swore them all to secrecy. They weren’t supposed to know anything about this. This was way before the transistor.
There’s a statement that was made by somebody else in the lab, Gordon Teal, and he said that there were few even within the Labs who knew about it for some months. It was a decision of management to maintain close security for a period, which is what we’re just saying, “In the interim period only those who were assigned to the project by research management were told the facts concerning the research.”
But those assigned included a large number of people besides us, such as H. S. Black…
…who was producing the point contacts…
He was measuring noise, frequency response and other things.
Right. And Teal’s point was that he had some suggestions on the single crystal, and that this might be very important in the transistor work, as important to the transistor as removing the last traces of gas from the vacuum tube, and he said in spite of his acceptable analogous reasoning, support was difficult to obtain.
I know this.
Now, what’s the story? What does he mean by that, about support being difficult to obtain?
This is a public talk Teal gave on this.
Teal had a habit of being very cautious about any ideas he had, and sensitive about other people taking his ideas. Teal’s a friend of mine, you understand, but this is true. Except for this characteristic, Teal might have been in this original group that was organized after the war. Shockley thought we could cut out single crystals. He didn’t see the importance of the single crystal. And I think this is one of the instances in the Labs that I mentioned before, in which Teal and Little, just by main strength and stubbornness, went to work and they grew single crystals. Schaff and Schumacher, and Schaff and Theuerer, in the metallurgical department, it was really in their area to do this, as an improvement of material. They didn’t want Teal in their group. In fact, he had a hard time finding a position after World War II. Nobody particularly wanted him. But after he grew the single crystals, they wanted him back in the group.
You mean in the metallurgical group?
This had nothing to do with the quality of his work, it was a question of his style and interactions with other people that were difficult?
Yes. And it is, in my estimation, one of the reasons he went to Texas Instrument.
He eventually got recognition from Bell Labs…
…oh yes, oh yes. It’s just an illustration that it takes many types of people to do things, and different people do things different ways.
What was Shockley’s response when he was informed of what you and John Bardeen had accomplished? First of all, how up to date was he on the day to day work? How closely was he really following it?
Oh, he went off by himself and worked at home, and in a way ceased being a member of the research team. The book is the result of this.
His book, you mean.
But when did that happen, after this?
I really don’t know exactly.
But between December 23rd and July you mean, or after that?
I don’t know whether this ought to be on the record, but he called both Bardeen and I in separately, shortly after the demonstration, and told us that sometimes the people who do the work don’t get the credit for it. He thought then that he could write a patent, starting with the field effect, on the whole damn thing, to include this. It’s only after it was found that there was a patent, by Lilienfeld, that read on the field effect, that this was abandoned. I don’t think this would have gotten through the Patent Department. Before our patent was written, Bardeen and I were both privately questioned as to the other’s participation. Both of us said that it was a joint project.
Do you recall when you first gave him the news? Before the demonstration, you must have been aware of what would happen.
Oh, he knew.
What was his response then, do you recall?
Publicly, nothing particular. The only thing was the private chat he had separately with Bardeen and me. I told him, “Oh hell, Shockley, there’s enough glory in this for everybody.”
That is the way the subsequent publicity on the work came through — that he was the head of the group, his name was always associated with it.
Orders came down the line, presumably because he had contact with Kelly, that no pictures be taken of Bardeen and I without he being present. Without his presence, I should say, for good English.
I think the language was pretty clear even in the press releases, or at least in the other statements — I mean, the patent, Western Electric’s I guess, on behalf of — they filed it, but it was your name and Bardeen’s name that was mentioned, that was all.
That was all. Well, read Gorton’s story — two years after — in Gorton’s story, the junction transistor is not mentioned.
All right, then we come to a whole…
If you notice, in that story, there’s a flysheet on top of it that we all agreed, including Shockley, at that time, to this story.
Let’s talk about that same six month period. It must have been difficult to keep this news from colleagues in the field outside of Bell Labs who were doing related work. You were saying something about the Purdue people, for example.
Yes. Well, we were concerned, in the group, over the fact that breakthroughs of this kind have a very definite habit of occurring in two places simultaneously, and we were very alert to the possibility that there might be another group that might have done this. We gave a demonstration to the military a week before the press, with careful thought that we were not going to raise the question of secrecy classification to them. We realized, if we asked whether it should be, that we’d be in trouble, because they would then have to respond to that. Our thought was that if we didn’t ask, and put them on the spot of coming out and saying so themselves, we would have no trouble. We told them that this was a demonstration that was going to be released to the press the next week. And they had to take the action if they wanted to stop it. And then, almost immediately, there was a statement came through to us that they had somebody working in one of their labs that also had something like this, and Bown and Shockley went down to the Washington area during that week, on a hurried trip, and had a session with the military management and a man. And Shockley very quickly proved to this man that he didn’t have anything like it. They had wanted to join in the press announcement.
What was the motivation in going to the military anyway? Was this supported in any way by military funds?
We felt that we had the obligation to inform the Defense Department of what we were about, because we knew that it had military implications. That is, we thought this was the duty of the Labs.
In this case, there was no support of any of this work by the military?
So there was no obligation as a contractee.
No, but Bell Laboratories has always had a feeling of responsibility to the defense of the country. Well, it came out of World War I.
They could have picked it up of course a week later from the press. They would have had as much information.
Yes. But then they could have yelled.
How come they weren’t told…
So this was a matter of courtesy.
A matter of courtesy, and also a matter of getting ourselves in a position where this would not be classified. We were very sure that we did not want that.
You mean, by giving them the choice and hoping they wouldn’t take it, then you’d have settled that problem.
Well, by putting them in a position where they had to raise this question.
During that one week period, they had exclusive information on it, besides the people who were directly involved at the Lab. They were the only group outside the Lab.
They were the only group outside of the Lab.
It would have put the responsibility on them, if they wanted to do anything about it.
Well, I’m sure we’d have tried to fight it, but the point is that if we asked the question, then they’d be afraid to say no. See, we’d put the responsibility on them. We presented it to them that we were going to release this to the world, and put them in the position of having the responsibility of saying no. This was thought out very thoroughly. The tendency would have been for the military to say, yes, it should be classified, if we’d asked the question.
The reality is that so many people were working on it that it was on the brink of being published then, anyway. On that point, you mentioned going to a scientific meeting, I guess APS meeting, I forget who the individuals were, — the paper…
It was either Bray or Benzer, who were then working very closely together at Purdue, under Lark-Horovitz. This was not the first abstract they’d presented on this phenomenon. We knew that they’d been working on this problem. The problem was that you could predict what the limiting forward resistance would be of a point contact rectifier, if you knew the specific resistance of the material on which you put the point. You could calculate the resistance, the spreading resistance from the point. It’s a very simple physical formula, well known. One that Bethe used and one that Becker and I had used before the war. It’s in any book on electromagnetism. And they were confused. They had discovered that the resistance in the forward direction on a high back voltage germanium point contact rectifier was less than the spreading resistance, and they were doing experiments to try to understand this. And one of the two gave an abstract at the American Physical Society. It was in the Bulletin. We could see it beforehand. On their work on this resistance. And of course, we attended the paper — this was the January meeting 1928 in New York. And of course we said nothing. We didn’t enter the discussion.
“We” means you, Bardeen, and…?
And Shockley knew. I don’t know whether he was there. We had visited Purdue, and I knew both Benzer and Bray very well. They were friends, scientific friends, and afterwards, whichever one it was, I met in the hall, and he wanted to talk about his paper, and I let him talk. Until he said, “You know, think if somebody put another point contact down on the surface, close to this point, and measured the distribution of potential around the point, then we might be able to understand what this is about.” And I couldn’t resist saying, “Yes, I think maybe that would be a very good experiment.” And walked away. Bardeen and I had already done this experiment. I probably shouldn’t have even said that. But that was the end of that discussion.
When was the next time you heard from them?
Well, the next time I heard from them was when Horovitz was invited to our press conference. We invited a large group of people to the press conference, and Benzer was on vacation in Long Island, and Lark-Horovitz told Benzer to be his delegate. When we walked into the auditorium at Murray Hill, the one that Fletcher and colleagues had designed, a good acoustic hall, and Benzer saw me, he said, “What’s this all about? We had some ideas about this.” I said, “Well, Benzer, I don’t want to spoil the story. You listen, and then you talk to me afterwards.” And afterwards Benzer said to me, “We had no idea of this.”
So there was certainly no ill feeling.
Well, they may have felt sorry that they hadn’t gotten onto it.
But they didn’t feel there was any borrowing of ideas…?
Did you get any kind of flack like that in that period? You mentioned the military.
There was one other incident, so far as I know. There was a press report, from France, that the French Telephone Company, they called it, P.T.T., also had this. And we investigated that, and found out that when our announcement was in the news, somebody in the French Parliament asked why the P.T.T. hadn’t done something like this, and the political answer came back up the line: Yes, they had. And through Aigrain and Dugas — whom I first met when Mott was at Pittsburgh, a session that I attended, because he was a solid state theorist, and they were there at that time, and we were friends with them — through them, Shockley and I got permission, through a connection at P.T.T. to visit the people who presumably were the ones that were supposed to have done this. They were H. F. Matare and Welker, Welker later did the 3-5 work. In an organization that was under contract to the P.T.T., to work in the semi-conductor field. And when we arrived in France, either Aigrain or Dugas told us privately that these two men had no claim to have done this, scientifically, and they would be glad to see us, but they did not wish to discuss this question. So we spent the afternoon. After our luncheon with the P.T.T., with only wine and no water to drink. A big luncheon. I couldn’t find a drinking fountain. I can remember this conference. Shockley of course was going great guns, and I was looking around, finally I got somebody to bring me in a glass of water. I thought I was going to die if I didn’t get some water. So we had a very interesting discussion of our work and so on and so forth, and that was that. We got some so-called devices — they gave them a different name — through some channel, that were being made in France, that presumably would do this. And we measured them and they were just junk.
That opens up a whole post—transistor story. It’s 12 approximately — shall we break and get ready for lunch?
I think, probably.
We have enough that I would like to continue on later.
We could do it tonight, if nothing else… I’d like you to see Larry Dodd, before we go to dinner, and see his setup. We could do that in 10 or 15 minutes after lunch.
Then maybe we’d have time for another half hour here.
We’re resuming now after a break for lunch and a trip to the library and the archives. We’ve agreed that we’d talk about the period just after the transistor. Actually our last discussion was about this six month period of keeping thin under wraps, completing the work somewhat, and about the public announcement that was made, and about the possible challenges on the transistor –- priority — and what did come up, how they were taken care of. I notice that in your files, there was a list of more than 300 individual transistors given out in the months following, well, not months, up through maybe 1950 — to various individuals and institutions requesting them. Was there any policy on that, in terms of who would get them and for what purposes and so forth?
Many people wrote to me, and probably many people wrote to Bardeen and Shockley, and there was a setup in the Bell Labs where, with due consideration, these were furnished, I think more or less free of charge, to universities.
For their own analysis and demonstration?
It would have been easy enough for any other company in the field to make their own, since the literature on it was published, or would that depend on the availability of materials?
Well, our laboratory also was setup for licensees, a program in which we from time to time brought them in and had big conferences, telling them the art and technology and the developments that went on. Any licensee was privileged to have this information. Also, early in the game, a matter of interest, it’s probably been mentioned, is that the ATT system agreed that there would be no license fees for transistors used in hearing aids, in honor of Alexander Graham Bell, who worked with hearing problems.
Then you got into cross-licensing arrangements I guess with companies like Phillips…
Yes, and the Japanese, RCA, General Electric.
This implies a big change for Bell Labs in terms of its business position, stature, and also its ability to have this device, which then became central to everything else in communications. What effect did the discovery of the transistor have on your working relationships and your environment at Bell Labs?
Well, the biggest effect on me was that there were demands for talks, which we distributed among us, and requests to appear at certain conferences and so on. We continued the work Bardeen and I were doing until we published the main article in the Physical Review. But I got into a situation where I was doing very little productive research. Oh, I went on and did things — the point contact transistor was a three-dimensional problem, which is not an easy problem, and one of the ideas, and I think maybe Shockley had a great deal to do with initiating this, was to linearize the problem. I made some samples, I had some samples of germanium made in a long strip, with the point contact, to inject minority carriers, on one end. We tried to make transistors of this kind and we did a lot of calculation, from this interpretation of our results, how far these minority carriers would drift.
The real work of this was done with Shockley and J. R. Haynes, in which they actually measured the pulse as it went down, the injected pulse, and as a result of this, measured the actual mobility. Up to this point, the only mobility known was the mobility you got out of the Hall Effect, and there were differences between these two. The effective mass of the charges came out of this. But don’t know as I participated in that, though the thing I did participate in was in measuring the lifetime of the injected carriers. We made great big single crystals, and there was a period we went through, making higher and higher lifetime material by careful treatment and purity control. The lifetime was a structure-sensitive quantity, even over and above the impurities, in the single crystals, how well you made them and so on and so forth. The lack of lifetime was due to other impurities, such as copper, interstitial copper in the lattice, not substitutional impurities. And I think I once had a large chunk of germanium single crystal. I forget, I’d have to look up in my notebook to see, but we measured lifetimes of something like a thousand seconds, in some of this material.
This was with the same group, or were you now working…
I’m working still in the same group.
Did the composition change at all?
Well, Gibney had left and Sparks joined the group.
Why did Gibney leave?
Gibney left because of health conditions in his family. His wife had asthma, and he’d already gotten an offer to go to Los Alamos, before the transistor, and he had, about that time, accepted, and he went on out to Los Alamos. Because of personal reactions that came in, when Sparks came in as his substitute, Sparks worked solely with Shockley. Sparks made the first PNP. He was not a member of the group the way Gibney was. The group began to break up, because Shockley went off by himself and did his work and so on and so forth. I guess Bardeen had finally gone on to working on superconductivity, and the group collapsed, somewhat. Bardeen kind of left the transistor work. In fact, he was kind of invited to, by Shockley. And Bardeen finally went to Illinois, and I was left, with Shockley. You might as well write this down, record this. The time came where I decided that the Bell Labs didn’t own my soul, and I went to Jim Fisk, who was then executive director of research, he’d moved up to replace Bown, and told him my problem.
I was no longer reporting to Shockley. And it became obvious in my discussion with him that he did not have the authority to do anything about this. So I kind of turned on my heel and said, “I guess I’ll have to go to M. J. Kelly,” who was then executive vice president. And I said, “Do you want to call him, or do you want me to call him?” He said, “You’d better call him.” So I called him, and he very thoughtfully arranged for a session with him at his home in Jersey. I was glad of this, because if we were in his office, his phone would ring all the time. You know. I went to his home. He’s a tough customer. I stated my case, and pretty thoroughly knocked me down on every question I raised. So finally, without thinking of its impact, I inadvertently said to him, that John Bardeen and I knew when Shockley invented the PNP transistor. This immediately got to M. J. Kelly, because he knew at that time — we were still not sure about challenges to the patents, on the two patents which were issued, first to Bardeen and I, and then written by a different patent attorney in different language, a PNP junction, to Shockley.
If the PNP junction patent had been written by some other concern, we’d have fought it, on the basis that there were claims in Bardeen’s and my patent that would cover PNP. I have here somewhere in my files a copy of something that Shockley worked out on the 25th anniversary, showing the various patents, and so forth, very interesting now… Anyway, Kelly realized that neither Bardeen nor I, if we ever went on the stand in a patent fight, would lie about what we knew. This changed his whole attitude. And after that, my position in the Laboratories was a little bit more satisfactory. I felt a little freer and so on and so forth. Then I went to Harvard, at Van Vleck’s request, which I think was ‘53, wasn’t it? And when I came back, I wanted to start research. Van Vleck by the way had mentioned to me that he had a student who was now at the Bell Labs that he kind of wanted me to look up when I got back. He thought maybe he wasn’t in the right part of the Bell Labs. I ought to investigate it. When I got back, I decided I wanted to start work on understanding the surface of germanium. And my idea was to use a germanium crystal in contact with an electrolyte, and to use pulse methods of studying the capacity as a function of potential. I went all around — I wanted another man to work with me. I like to work with at least one other individual. I feel more productive that way. You’ll notice, a major portion of anything I’ve published has been in cooperation with some one other individual. Shockley was still in a position of having considerable influence on what kind of projects were worth doing in this area, and I got no response.
I wanted a chemist to work with me on this, because it was partly a chemical problem. I got no enthusiasm anywhere, till somebody quietly said, “Why don’t you go to Jack Morton?” who then had the control of all the development on the transistor. So I went to him and he said, “That’s just what I want somebody to do, and besides I’ve got the man that will work with you.” And the man turned out to be C.G.B. Garrett... (off tape) So I went to Jack Morton and he said, “That’s just the kind of studies I want made.” The surface problem was then the biggest problem in the whole development of it. And an understanding of the surface was important, and he said, “I’ve got just the man to work with you,” and it turned out to be C. G. B. Garrett, who was the man Van Vleck had told me about, and whom I hadn’t had an opportunity to do anything about. He said, “You can work together, but he’ll stay in my department.”
After your conversation with Kelly, when you conferred with Kelly, you said at that time you were no longer reporting to Shockley.
No, before Bardeen left, I told you that one Friday, we walked into Fisk’s office, when he was just our immediate supervisor, over Shockley, and told him that we did not wish to report to Shockley any longer. And Monday morning, we weren’t reporting to him. It was about, oh, six months. Kittell, also for reasons of Shockley, left for California. In fact I held a farewell party for them, and in inviting people to the farewell party, I kind of worked up the organization chart of the Bell Labs. And nobody refused me. So I went clear to M. J. Kelly, and he didn’t refuse either. Remind me to tell you something later about this that I don’t want to put on tape.
Just getting the chronology — Bardeen left, what year was this?
You’d have to look that up. [before 1953]
Bardeen left because he preferred an academic environment, or because he felt it was no longer comfortable for him to continue at Bell Labs, or both?
Bardeen was fed up with the Bell Labs, with a particular person at the Bell Labs, and he made up his mind to do this about a year before he left. I told Bown almost two years before he left that this was going to happen if something wasn’t done. And Bown didn’t pay any attention to it. But when Bardeen made up his mind, tremendous pressure was brought on him to stay. And when Bardeen makes up his mind, there is no use doing anything about it. It is too late.
One wonders, why Illinois of all possible places.
He was made an offer that he liked there. Bardeen and I continued on, finishing up certain research on germanium that we worked together on. [Bell Syst. Tech. J. 32 1 1953.] Anyway, this was published after he left.
So then after your conversation with Merton Kelly, were you and Bardeen part of some specific group, if you were no longer reporting to Shockley? What were you, your own group or what?
I’d have to look up an organization chart to see who we were under. I ended up being under McKay for a while. This was still later though, I think.
But after that conversation with Kelly, where he got the point that you really should be taken seriously, you said things were better. Did that mean that…?
Well, I felt that at least Kelly knew how I felt, right at the top of the organization, and I felt much more secure in my position. There were things that were said and done that I didn’t like. I think this was probably the beginning of Kelly’s realizing that maybe his, all his information came from Bill Shockley.
Prior to that, you had no real direct channel of communication with him.
I hadn’t tried to. I hadn’t thought, up to this time, it necessary to go this far in the organization. I went to Fisk first.
Once you’d gotten it off your chest, even that was a relief.
Oh yes. Well, Garrett and I started to work, as I told you. I think I already told it on this tape. We worked for a year and a half without any results, and then one day we had everything right. And wrote the paper. We had some fun, because the first paper we wanted to write, we were immediately faced with the question, which organization channel does it go up, through Morton’s or through the research department? And when I found out that there was going to be an attempt made for it to go up both channels for clearance or publication, I just raised hell. I said, “We’ll trade, this one goes up through physics, the next one we publish can go up through yours, but we’re not going to go through the nonsense of putting it up through two organization channels.” I think it’s of interest here to mention the fact that lots of young men came in, and in these days have some choice of where they want to work, and it was easy to get them to work for Bill, to get them to work with him.
One young man Bill put to work on the idea of cyclotron resonance, and by the time he sets up the apparatus, and the first results he gets were negative, as usual in an experiment. Remember, Bill’s a solid state theoretician, an excellent one, Bill would have another idea. He’d want the experimental man to switch to this other idea. And when I was at Harvard, Kittell wrote to me for some silicon samples, for some purpose, and having these silicon samples, he got the idea of also trying for cyclotron resonance on them. The man working on this experiment only went down to the temperature that Shockley said the cyclotron resonance ought to occur. Kittell went to a lower temperature, and got cyclotron resonance, right out under the nose of the Bell Labs. And he insisted on putting my name on the paper. I rebelled at this. Because I had just helped him with a sample. A young man in the metallurgical department by the name of Pfann had the idea of zone melting. Shockley didn’t think that was essential to the program. And suggested that Pfann should do something else. And he ran bang into another character, someone like I am, Schumacher, who was head of the metallurgical department, and he just told Shockley to keep his blankety blank fingers out of his department. And of course Pfann went on to do his zone refining and is now a very famous individual for all aspects of this work. Shockley admits some of this in the paper he’s given me to look over, that he’s writing, possibly for publication. I haven’t heard since I criticized it. The paper was good, and he admits things like this, now.
Was he getting in a somewhat isolated position within the laboratories?
Well, this is another story. You want to jump to that now, or want to keep it? All right. When I was approaching retirement, I was very carefully asked if I would object if Shockley was made a consultant of the Bell Labs. And I laughed at them and said, “I know, you want a Nobel Prize winner around for a little longer, a Shockley is a very very intelligent solid state theoreticker, if you want to do this, OK.” In fact, this was partly Jack Morton’s idea, because Shockley, in his attempt to go out and make a million dollars, on Shockley’s Semi Conductor company, lost control of it. It left him without much income, because he was taken out of that, you know. The company was sold. I was told very carefully, he was going to be brought back into the laboratories on the basis that anybody in the laboratories that wished to could consult with him, but he was going to have no authority to dictate.
When did he leave the Bell Laboratories?
I’d have to look up dates for this. [Between ‘54 and ‘57]
Was there a deterioration of your personal relationship on a day to day basis during that period, from the transistor time on?
Oh, I’ve always been polite to him. Spoke to him. I don’t relish, today, getting into an argument with him, especially on his genetic ideas. I’m surprised a little bit, though, in this article, that he is beginning to realize and admit some of the things he did.
Getting back now to the other aspect of the transistor work, besides the effects — this was a turning point in Bell Lab’s history. How would you describe the morale, beyond the group itself, after the transistor, the whole spirit of the Lab? Was there any noticeable change?
No, no, I think the Bell Labs went on quite well. It was probably impossible for that group to be maintained together as a research group. But I was saddened at the fact that it should break up, as a group of cooperating scientists. It was a powerful group.
The breakup started to take place just as soon as the work was done.
Well, the first step was when Sparks came in to replace Gibney. He was really not a member of the group.
When was this?
It must have been in the first part of ‘48.
Even before the announcement.
Yes. You notice in here that you find Sparks’ name on the second one of these memoranda.
It’s the April one.
So that group existed really, its active period was about three years at the most.
I can tell one story, if I haven’t told it in the other transcript, about the quality of this group, how it acted. There was a man out in California that was publishing papers on the fact that NH3 was a superconductor at high temperatures, relatively high temperatures. And of course, any superconductor at high temperatures is of interest to the communications industry, if it’s real. And this problem was put to this group — maybe the larger part of the group, the whole solid state group, not just the transistor group. Well, we called a meeting. They had one person report on what we knew about these papers. We decided to let this jell, and we called another meeting very shortly afterwards, in which we discussed the whole question, and decided that somebody ought to go out to visit this man. As usual in the group, the proper man to go was always obvious, and it was G. L. Pearson. This I believe was at Stanford and Pearson had a Masters degree from Stanford. He went out, visited the man in his laboratory, came back, described all the experiments he’d seen, etc. Then the question was raised in the group, what’s the simplest experiment that will be definitive here? See, this man had all sorts of complicated experiments, that showed a great reduction in resistivity. And the answer was to test the magnetic susceptibility of this material.
If there was a transition to superconductivity the magnetic susceptibility should change in a certain way. And who was going to do the experiment? Well, Pearson. So they fixed up a big sample of it, and let it cool down — in a system that measured the magnetic susceptibility, and the magnetic susceptibility results showed it wasn’t a superconductor, and this was published in a letter to the editor of the Physical Review, and that was the end of that one. But the whole point is, we didn’t do any of the complicated experiments. The group asked themselves, what is the simplest experiment that will be definitive? And we did it. And at that time, there was very little question as to who was the right one in the group to undertake this. And the same way after Bardeen indoctrinated us on the effect of surface states, and as a result of this discussion in the group there were two suggestions made, one by Shockley and one by me, which are a matter of record, of course, and I was the surface expert, it was obvious that it was up to me to do the experiment. But you got ideas, in this group, which you wouldn’t have gotten without the stimulation.
Maybe something like that does have another half life, that’s so effective, so intensive —
Yes, and as things come in, personal things like Gibney’s situation, why, —
When the transistor was finally real, in a way the mission of the group, which had been defined somewhat differently, had been accomplished. What was involved then in the transistor was a great deal of development after that point.
Oh yes. Well, we did not have to do the development. This was immediately transferred to a development group.
And you were free to continue…
Many people in the world today think I’m a top technologist, transistor technology, and I know nothing about it. I just have the overall picture.
About how long was it after the July ‘48 public announcement that Shockley had, you know, come into the picture on his part of the transistor work?
As I said before, he immediately started working on his own, beginning early in 1948. He developed the PNP concept then. The PNP transistor idea, at first had to wait on single crystals, Teal’s work. And Morgan Sparks was the one that undertook to dope a single crystal, starting out, first one type, and then, a little over-compensating of that impurity, and then going back to more of the other, original type, so that you had N P N junction’s. With the P type, if it was that way, I don’t know whether it was P N P or vice versa, but with the middle type very narrow in the middle of the crystal, and they cut the device out of a single crystal.
You say when the idea came in his mind, which was when, right after?
After he went home and went to work.
Which was practically when?
This was between January 1st, 19148, and somewhere after the announcement.
The group almost ceased to exist in the formal sense by December 23rd, 1947. Well, that isn’t quite getting the chronology right, I understand…
I wanted to put in just a remark here, you can make a note of it, that some of the anecdotes that I’ve been telling — some of which I’ve been a little bit afraid to tell on this tape — were actually told at the IEEE Conference on March 23, 1973 and are on a tape made at the meeting which is available.
Just in public, so we shouldn’t refer to...
I told you that I had planned to ask you a lot about the Nobel Prize, but then last night I had a chance to see most of the account, 62 pages, that you’d prepared on the saga of your trip to Stockholm. I’m reduced to one or two questions about it. When did you write that saga, immediately?
Upon my return, yes. Well, I was encouraged because Polykarp Kusch had given me his story, and I felt it somewhat my duty to share my experience with my colleagues in the labs and outside of the labs. I have a list in there, up to a certain period, of everybody I sent a copy to.
Yes, I know where that is now. Well, because the details are so well described in such good narrative and sequence, I don’t want to rehash that. It was fresh in your mind then. But there’s one thing, you start off by saying that people asked about your reactions, and also about, did you expect this? And you said there that it had come to your attention that you had been nominated, that even when you did the work, you had a feeling that this may be of Nobel Prize stature. That’s what I want to explore just a little bit. What’s the first time the thought entered your mind, or it was mentioned to you, that that work that you were doing in this group may lead to a Nobel Prize?
I cannot definitely say when. We knew immediately, the whole group, of the tremendous importance of the transistor. We knew the importance even of the amplifier in principle and the field effect. And we also knew of the importance of this. While we may not have predicted in our minds the rate of growth, and we had no idea of some of the developments, integrated circuits, etc., we knew its importance. We knew its importance especially in connection with the point contact, because the point contact transistor, as I’ve told in “The Pierce Story of the Naming” was a dual of a vacuum tube. The point contact transistor was short-circuit unstable. The vacuum tube is open-circuit unstable. If you leave the grid floating, it will sing. If you shorted a point contact transistor, it would oscillate and burn up. With power on it, of course. This was the chief difficulty that the electronic engineers had. They’d burn up transistors as fast as they’d get them, because they’d try to make the impedance high, in the passband they wanted, and short it outside, from their experience with vacuum tubes. And because it was not just a substitute for the vacuum tube, but it was a circuit element with other properties, we felt, early in the game, that it was more than just a substitute for a vacuum tube. And of course, we knew that it was a landmark in understanding of solid state. We also knew the need for single crystals, etc. of high purity. The phase boundary in a single crystal of a covalently bonded lattice was another surface model, like Langmuir’s clean tungsten in high vacuum. It was on the other end of the spectrum. It enclosed the more complicated surfaces. All these techniques went immediately all over into solid state.
You said that was all clear to those of you who had participated.
Very rapidly clear.
So what you’re saying is that the idea of personal recognition for it was a consequence, I’m assuming you’re saying this was a consequence of your recognizing the very large significance.
Well, there’s my statement in the saga that any physicist is aware of the relationship between the important work that he’s done and a possible Nobel Prize. I covered that, I thought, pretty thoroughly in my story.
The one that I gave you a copy of, that you read. The saga. I should say, the Academy of Science of Sweden, which covers the prizes in chemistry and physics has a rule that any Nobel Prize winner in this area is offered an opportunity to nominate, every year. And they also send requests for nominations to people in fields that may not have been represented. I won’t mention names, but I know one instance in which the man who nominated both Bardeen and I told me, and I know another instance in which we were told, that we’d been nominated, the three of us together. And this man with considerable thought, and he was not a friend particularly of Shockley’s, in fact he was on the opposite side, decided that this was the only real way to get the prize for any of us — that Shockley should be included. And I think he should have been. And I learned afterwards that Davisson also nominated us. But then of course, the Bell Labs being the kind of organization it is, and the efforts of the press to get inside information that’s supposed to be private out of the Academy led to fresh information. We learned that we were running high in voting, two years before we got it. Well, I told all this.
Right, that plays hell on the nerves. So when the prize finally came with all three of you, it was no surprise that the three of you were together. Or was it?
We were pre-warned that we were going to get it, that this news was coming, the year we got it.
I see, that it would be all three. Well, you told the rest of that story. We don’t have too much time and we should have a break — so let me, if you have more patience —
Let’s talk for a minute about the ultimate decision to leave Bell. Was it the normal time for retirement? What year?
I told President Perry of Whitman, in fact I told some men at the University of Colorado once, that my period of retirement was approaching, and that what I wanted to do was to associate myself part time with some not too large college, and maybe do a little teaching on the undergraduate level. And I told this story, without thinking about Whitman too much, to President Perry. Also involved in this was the fact that my first wife died in the spring of ‘57, from cancer. And that at an alumni meeting shortly thereafter in New York, I was introduced to my present wife, with the comment that “You’re both in the same position.” She had just lost her husband from cancer. Her husband was managing editor of the Walla Walla Union Bulletin. And I went West as much as I could, and I was elected to the board of overseers of Whitman College. I came in contact with President Perry, and some time, somewhere in here, at cocktails or something, the question came up and I told him what I wanted. He said, “Why don’t you do it at Whitman?” And my wife came out of course to New Jersey.
The oldest daughter was here in college, but she brought the younger daughter and the boy with her. There turned out to be problems with raising, particularly the daughter, in this region where she didn’t know the families of the boys that the daughter had dates with. And she finally made the decision that she was going to have to take them back to Walla Walla. And I acquiesced in this. So Perry said, “Why don’t you do this at Whitman College?” So an arrangement was set up as a result of this talk with him, that I come out here a week a month during the academic year. I would fly out on a Friday, go down to Newark, get on a plane, get to Seattle, still daylight, stay overnight with my sister, take an early morning plane to Walla Walla in time to get my golf game in on Saturday afternoon, play Saturday and Sunday. Monday, I would have the class in, see what they’d done on the last experiment, tell them what the next experiment was. I came out here and looked over the available apparatus and designed my experiments, you know, with the apparatus, as I’ve already told you — not on tape but when I showed you the notes on the experiments —
On Understanding Science?
No, this was the senior laboratory, for senior physics majors. And after doing this for about two years, I think it was, I got to a place where I was 60 years old. And I had already discussed this other course with the scientists here, thinking of Professor Brown’s course on geology and astronomy, when I was a student at Whitman, and there was no course here like it. There was astronomy and there was geology, but there was no “Understanding Science” course. We discussed how we could implement this, and decided we’d have to have a summer session and give it to ourselves. I went back to the laboratory and some place picked up Weiskopf book, “Knowledge and Wonder,” read it through almost in one sitting and thought, “I can teach this course out of this book as a text.” “I don’t need any help.” So, I went to the Labs and asked for retirement. They said, “Oh no, we don’t want you to do that, you’ve got five years more to go.” This may have been a little bit beyond 60. “We’ll set it up so you can spend the academic year out there. Everything you do except in your teaching at Whitman College will be at our responsibility, your travel, your talks, attending conferences. We will pay you approximately half the salary you’re getting.” So I immediately raised the question, retirement is based upon your last ten years of salary, as to how this was going to affect — I didn’t want my retirement income to be any less than if I retired right then. And they said, “Oh, your nominal salary will be your full salary, but we’ll only pay the half part and the expenses.” They even raised my salary during this period! So I came out here, spent the winter out here. I was on the Naval Research Advisory Committee, the Defense Science Board, and these boards of course financed all my travel — whenever I went East to one of these meetings, I’d drop in at the Bell Laboratories, at my office, spend maybe a couple of days there, come on back out here. Summers, I’d go there. Or if there was a semi-conductor conference, I was the chairman of the semi-conductivity commission of IUPAP — so Bell Labs supported that. Well, when ‘67 came, I was 65, and through with the Bell Labs.
So that was the end of it.
Did you manage to get anything done in the few days that you showed up there during that interim period?
Oh, I didn’t do any research at that time. Well, I was getting interested in biological research then. As a result of my son’s operation, open heart, by a doctor in Seattle. He took me into his lab after Bill was safe, wanted to show me all about blood, arteries, how they worked. The one thing I could get out of it (their language was different from mine), one of the major problems was clotting where they used vascular prostheses and I said, “This looks to me like an electrolytic problem.” I said, “Has anybody ever measured the potential, on the hydrogen scale, at which red blood corpuscles will stick to various surfaces?” The doctor’s name was Lester B. Sauvage. He said, “Well, I don’t know, but there’s a friend of mine back at Downstate Medical on Long Island who will know the answer to this, and we’re going to be out there next fall — will you come and have dinner with us?” Well, I was still in Bell Labs. I said, “Fine.” Once a month out here. So I said, “Yes.” I was then working with P. J. Boddy — still on germanium surfaces and electrolytes. So I went over to dinner and listened to him.
This Dr. Sawyer was hell on wheels and if you cornered him with a question, he’d change the subject and go on talking about other things. And it looked to me like he was measuring primarily electrical transport properties of blood — red corpuscles are charged in the blood stream — well, I came back and talked to Phil Boddy, who was a physical chemist, and we had as a result of this dinner, a conference with him. And some place in this I said to him, “It seems to me, the first thing to do is not any transient phenomena in this area, but steady state, equilibrium phenomena,” and I said, “How about measuring the potential?” So we told him what we wanted to measure and he said, “I’ll do that. I’ll measure that and tell you in about a month. Well, he didn’t tell us in a month. He waited finally till he got an assistant for the summer, and he set up an apparatus, and he came back over to the labs. We were busy on our own work, and so we invited him to come over. He told us what h was doing, and we told him he wasn’t doing it right, told him how we wanted him to do it. Finally he said, “Come on over, we’re doing it right now.” He had a little setup with salt bridges and the proper potentials, with ability to change the potential with a platinum contact that went down into a little dish, and to get fresh blood that would still clot, he just slit his thumb, his finger, like you do when you take a blood sample, put a few drops of blood down in a Krebs solution.
Then he’d look in the microscope, and he could vary the potential between the platinum and the solution, and you could see that if the potential was one way, the corpuscles would hit the platinum and bounce off at random, there was no current flowing, there was no obvious drift of the corpuscles, because in the saline solution that he had them in, the potential was practically zero. All the potential drop was at the platinum surface. But if you changed the potential the other way, then when they touched, they’d stick. So I said to him, “Here, you can see better than I.” It’s a little bit hard to see through a microscope — I said, “You’re used to looking at these things. You look in the microscope and I’ll vary the potential, and you tell me when they’re sticking and when they’re not sticking.” I varied the potential around and he said, “They’re sticking.” I said, “Just a minute while I read this potential.” And he started to look up, to look at the meter. I said, “No sir, you look in the microscope.” And I kept on doing this until I had readings above where they stick and below where they stick, and finally said, “OK, that’s it, I can average these, for the potential.” We had this whole series of papers on it.
I’ve seen some of them. So that whole new biological interest started.
Well, no, Garrett and I, and Phil Boddy and I both began to recognize that there was a close analogy between the electrolytic work we were doing on the germanium surface, and the nerve phenomena. We recognized that the nerve was also another active circuit element. Garrett went over to the public library and read up on all Hodgkins’ and Husley’s work, came back and reported to me on it. But we didn’t know how to get a biological surface that we could work with. We were busy doing our own work, and unless we could see any significant experiment, we didn’t see any reason for starting anything, until… (off tape)
All right, we’re in the middle of the biological study.
Well, anyway, somebody came across a little article in Progress of Surface Science by Rudin and Muller, which they called my attention to, in which there was a report that they could form, under saline solution, phospholipid bi-layers. Just like a soap layer, like on a ring that you dip in a soap solution and then lift out that collapses after water evaporates out, and you have a bi-layer, all the same molecules. And they had been able, by treating the bi-layers with certain proteins, to get action potentials. Qualitatively the same as the action potentials in nerves, or now known in all cell surfaces. Even simple cells, where you put micro-electrodes in them. So at my earliest opportunity I made a date with them and spent a day with them. And maybe spent another. At any rate, they invited me to come and spend some time in their laboratory. I came back here to Whitman and talked to Professor David Frasco, a chemist on the campus. He was interested. So I arranged, while I was still being financed by the Bell Labs, to go back there and spend a month with them in their laboratory. There was some research money that Frasco could use to go back there for two weeks to look at the chemistry. Went there with some of my constant current pulse apparatus, and my ability to measure capacities in phase boundaries of semi-conductors, which of course told the whole story at semi-conductors surfaces. But of course these films were only 75 angstroms thick — they were just a great big leaky capacity. And the capacity was only a function of the thickness of the film and its dielectric, and this didn’t vary. In fact the capacity was a constant of the system, so I didn’t find anything out. But we came back here, spent a summer, and a little research money and apparatus available here. We had some Keithley electrometers. That’s all you needed, two of them. And after the summer’s work, we had to ask ourselves, how are we going to finance this research program? I was very adverse to government contracts and all that red tape. I said, “Let’s go over to Battelle” — I knew Fawcett and “see if we can’t sell them the idea, and they can get the contract if they want and take care of the red tape.” Whitman College had no real mechanism for this. So Fawcett invited me over, after I told him this, and he had a group come in, biologists and others who might be interested, and we presented our story, and what we knew about Muller and Rudin’s work. The biologists kind of turned up their noses. They were all interested in genetics anyway. But there was a bio-chemist in the group by the name of Donald Kalkwarf and this intrigued him, and so he initiated a program, where we’re consultants. Dave gets a consultant fee. I told them they could give anything they thought I was worth to Whitman College. And we’ve been at it ever since. We’ve published two papers. They’re a little bit disgusted with us because we don’t publish more, but as long as I’m in the program, we’re not going to publish papers just for the sake of publishing.
You wait until you feel you have something significant.
Yes. What more questions have you got?
Well, just the final questions. We don’t have much time. Reflecting over your entire life, including student days and childhood and your work at Bell Labs and subsequent work, is there any period that you’d identify as being the most satisfying, personally pleasurable period?
Really all of it’s been very satisfying, even when I did what’s illustrated in the picture over your head — you can look at it afterwards. (Herding cattle starting in the summers when I was 14 years old. But I did not like following three horses and a harrow in the dust.) My chief impression is that I was always fortunate by being many times in the right place at the right time. There were steps in it when I didn’t know whether I was going to get to make the next step or not, in which I was worried for a while. My favorite statement is that I’ve been indeed fortunate to spend my life trying to understand how things happen. To be able to earn my living this way.
Fine. What about the things that you felt were most productive, in terms of satisfaction of knowing something specific that wasn’t known before? Is there anything like that? If you don’t mind talking…
I’m willing to continue talking for a while. I’m not sure I understand your question.
Well, what I mean is, let’s say within the scientific work itself, was there any particular — you did the surface work in the thirties, the transistor work — there are differences. One is a thing which made a large impact on the scientific community. The other could be, it needn’t be different from the other, but something that gave you personal satisfaction in the day to day work.
Well, I would say this. Except for my Ph.D. thesis, which was published in The Physical Review, which I had a feeling was not a complete job, but was finally accepted by Tate, which I had to complete and labor with after I left Minnesota and could not do any more experimental work, I’m very proud of all the things that I’ve published scientifically in recognized scientific journals. I have a feeling that they were all sound scientific papers. Of course, the one big event, the transistor — I happened to be there at the right time doing the right experiments. About five experiments that I did for reasons other than what they amounted to. Very definitely. It was a great period. That whole period in the group was a great period, too. Oh well, the interaction with everybody I interacted with. Fisk, Shockley, Teal, Holden, A. H. White. My late wife and I used to introduce the new young men and their wives, kind of, to Bell Labs. We’d invite them out to dinner or something. Some of these were Pierce, McKay, Charlie Towns, Kittel — I’ll run out in a minute — well, Shockley and his wife and the Bardeens. We interacted socially with them. We became early friends of theirs when they first came to the Laboratories. Used to tell each one of them that if they kept their nose clean and did a reasonably good job, the only thing that they’d ever get fired for from the Bell Labs was fooling with the female help on the premises. This is a very definite fact. I know of instances where people left the Labs just like that, for that reason. It was an old tradition in the Telephone Company because the Telephone Company had so many girls on the exchange, and it became a tradition in the ATT system, you did not fool with the female help on the premises, period. It didn’t matter whether you were president, vice president or what you were, you went out, if that happened. (Mills who was in charge of hiring service personnel had a habit of picking good looking girls. Some of them became Bell Labs’ wives!
Were there instances you know, in the research group?
We had one man in our department that left very quickly, because he got a secretary pregnant.
Didn’t help her any.
Well, I don’t know what happened to her. But that was the way it was.
Well, I think that really winds it up pretty well, for what we’re trying to do. Sorry to intrude on our conversation with a camera, but I —
It’s perfectly all right.
Well, thank you very much. I think we did a fine job. When you get the transcript, you’ll have a chance to — you’ll probably add things, rather than take them out.
January, 1964; at American Institute of Physics.
Bardeen Shockley adn I also published on article on an experiment that Shockley and I had done on oxidation of copper using radioactive copper before World War II. J. Chem. Phys. 14, 714 1946.
Metal rich deposits occur first in the form of sulfides but when eroded off generally become oxided to some depth in the formation.
1 1/3 feet in diameter and 2 1/2 feet long. The memo on this is in my files.
I think not! On looking at copies of my note book the first time, the device that was reduced to practice on December 23, was put together on December 19th after the memo on December 17th was written!
Noted by Brattain later: Bardeen left before my conversation with Kelly!
Shockley's group was split off from S. O. Morgan. I was under Morgan.