Oral-History:Charles H. Townes (1992)

From ETHW

About Charles H. Townes

Charles Townes.jpg

Charles H. Townes, a 1964 Nobel prize recipient in physics, was a pioneer in the field of laser theory.  He received the IEEE Medal of Honor in 1967 "For his significant contributions in the field of quantum electronics which have led to the maser and the laser."

Educated at Furman University, Duke University (master's degree) and the California Institute of Technology (Ph.D. in physics, 1939), Townes became a research physicist at Bell Telephone Laboratories in New York, collaborating with a variety of research groups and developing a number of patents. Following his employment at Bell Labs (1939-1947), Townes joined the faculty of Columbia University, where he continued his research on the microwave spectroscope. The maser (an acronym for microwave amplification by stimulated emission of radiation) Townes developed at Columbia and patented in 1959 was the successful result of his search for a way to obtain stronger radiation with shorter wavelengths using a source other than a vacuum tube. In collaboration with his brother-in-law Arthur Schawlow, Townes developed a laser, which could operate at wavelengths a thousand times shorter than masers. Laser technology, from which neither Townes nor Schawlow profited, changed the sciences of optics and electronics.

This interview covers Townes's education and research. Townes describes the work environment at Bell Labs, with particular attention to the collaboration of engineers and physicists. Analyzing the influence of World War II on Bell Labs and on scientists' goals, Townes describes research in computers' use to guide anti-aircraft guns and in radar bombing systems. While narrating his research at Columbia and his involvement in the Institute of Radio Engineers and the IEEE, Townes analyzes the interaction of physicists and engineers in the fields of radio astronomy and microwave spectroscopy.

See also Charles Townes Oral History (1991) and From Radar Bombing Systems to the Maser: Charles Townes as Electrical Engineer.

About the Interview

Charles H. Townes: An Interview Conducted by Frederik Nebeker, IEEE History Center, Piscataway, NJ, USA, 14-15 September 1992

Interview #143 for the Center for the History of Electrical Engineering, the Institute of Electrical and Electronics Engineers, Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Charles H. Townes, an oral history conducted in 1992 by Frederik Nebeker, IEEE History Center, Piscataway, NJ, USA.

Interview

Studies at Cal Tech and work at Bell Labs on the eve of World War II

Nebeker:

This is Rik Nebeker on the 14th of September 1992. I'm in Charles Townes's office in Berkeley. It was in the early summer of '39 that you completed your Ph.D. at Cal Tech and were hired by Bell Labs I've heard the story of that exciting trip to Mexico that you made on the train fare that you were given. I wanted to jump to the end of that summer. One thing, is what was happening in Europe then. Do you remember, for example, the German-Soviet Nonaggression Pact in August?

Townes:

I certainly do, yes.

Nebeker:

Did that suggest to you that war was imminent?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_1.mp3)


Well, what I remember is a little bit in hindsight now, but my impression then was what I guess everybody knows now, that Hitler was trying to shore up his possible defenses and protect himself on that front so that he could pay attention to the rest of Europe I think that's what [Chuckling] we thought it was at the time. You may have seen from these other interviews that German students were at Cal Tech and we talked with them interestedly. Most of the German students sent over had to be good Nazis, at least thought to be good Nazis, though they weren't all; some of them were pretty skeptical. Others were very strongly Nazi. We had peace marches on campus against the military-industrial complex--although that phrase wasn't used at that time--but against industrialists who were trying to get us into the war so they could sell more stuff. That was the claim. I don't think those marches were taken very seriously, but they were there. We discussed the situation and felt that we were going to have to be drawn into it somehow. It seemed very likely. Roosevelt, of course, was making moves in that direction to help. So the war clouds were all there. Various people took them with various degree of seriousness. But I think many people were quite concerned about war.

Nebeker:

Were you worried that a war would disrupt your career plans?

Townes:

I was more worried about whether there would be a war and what the United States would do. Obviously my career plans would be involved, but I don't remember thinking in those terms particularly. I was just worried about the general situation.

Nebeker:

Do you recall when you first heard of the German invasion of Poland, the 1st of September?

Townes:

I don't recall just where I was.

Nebeker:

You must have on your way up to New York.

Townes:

That may well be. I think I was supposed to report on the 1st of September.

Nebeker:

I noticed that the inside cover of your first notebook at Bell Labs is dated the 12th of September, but the first entry is the 21st of September.

Research groups at Bell Labs, Townes's patents at Bell Labs

Townes:

Yes, well, when we got to Bell Labs, they had an interesting orientation process. Bell Labs, did a very good job in introducing people to the Labs. They took all the new people, about 12 or 15 around the Labs.

Nebeker:

Were these all Ph.D.'s?

Townes:

Yes, I believe they were all Ph.D.'s, including Jim Fisk [James B. Fisk] and Bill Baker [William O. Baker] who were classmates of mine. They both became heads of the Laboratory. Jim Fisk had taught for a few years in North Carolina, and then come to the Lab.

What the Labs did was to give us an educational briefing which probably lasted a full week. We were shown around the Labs. Buckley, who was then head of the Laboratory, talked with us about his own career and how he saw things and so on. Get all these young people and give them all a little perspective.

Nebeker:

Do you recall if he said anything about the possible impact of the war on Bell Labs? Bell Labs at the time was doing just a small amount of contract work for the military.

Townes:

Right. I don't remember his saying anything, but that doesn't mean he didn't. It was, a substantial talk. Informal, but very substantial. It had a lot of parts in it, so he might well have and unless it was something quite unusual, I would not have remembered that.

Nebeker:

That would be the first week or so of September?

Townes:

Probably, yes. That would be my guess. So that may well be why I didn't start writing anything for a while. My impression is that I was supposed to report on September 1 but that was just a round date that they had picked. We had that briefing. As you know from my other interview, Bell Laboratories was also particularly generous to me. They knew I wanted to do physics. Mervin Kelly's idea was to send me around to four different departments for three months each to get acquainted with the Laboratory. Then this would show me and show them what I was best suited for and what I should be doing.

Nebeker:

But there were more than four departments in physical research.

Townes:

Of course, there were. [Chuckling] Four was a sample, you see. I didn't know what four they were going to be, but they said four, and I said, "Well, that sounds great." As it turned out, the four that they picked were not all in physics at all. The first one was microwave tubes, which really was an engineering group. But it was a very advanced engineering group, doing fairly fundamental research work.

Nebeker:

This was Llewellyn's group?

Townes:

That's right, Llewellyn's. That was my first assignment. microwaves were new and interesting, and there were plenty of basic things there. So that's how I started working with that group. I was then housed in the Graybar Building which was space they had rented in the middle of lower Manhattan because the old West Street building was full. So I was there.

Nebeker:

Was it Llewellyn's group that was there?

Townes:

Yes, that's right.

Nebeker:

More than that?

Townes:

There were some other groups, but there were all kind of related. I think there was some microwave communication systems group or something like this. Llewellyn's group was not a large group.

Nebeker:

Here is a memorandum that came out from Llewellyn at the end of that month, the 25th of September. It's considering frequency modulation.

Townes:

Oh, yes. Well, these are all famous names I remember very well. [Chuckling] Chaffee, Dietzold--Bob Dietzold, a good friend of mine. K.C. Black. Les Peterson. I thought it was Les Peterson. There was a Les Peterson I remember very well. He was a very nice person, and I learned a lot from him.

Nebeker:

There's a schematic sketch on the very last page of this system. But I'm guessing that you may have come too late to be involved in this.

Townes:

Yeah, I was not involved in that.

Nebeker:

This was written up just a couple of weeks later. This is a later memo from Llewellyn's group.

Townes:

No, I was not directly involved in that. I was put in the group, and I'm not sure that I exactly had an assignment. I think the idea was for me to just get acquainted and work with people. If I saw something to do, I would do it. Otherwise I would talk to people and try to help people, and see what happened.

Nebeker:

How large was that group?

Townes:

I don't know the exact size. I interacted with about ten or twelve different people. It was largely done on the basis of whom I found interesting and whom I was sitting near. I would listen to what they were doing and try to learn about it generally. There were some other groups in the building. Chuck Elmendorf, for example, was in a separate group.

Nebeker:

This was in Department 325, Circuit Research, with R.C. Matthies as director. Do you recall him?

Townes:

No, I don't remember him.

Nebeker:

But you were in 325D which was the designation of this.

Townes:

By now you know much more about it than I do. [Chuckling]

Nebeker:

Evidently what you were working on in September, October, November was a project, or, as they say, Case 37010. It's listed here on these pages. The first 20 pages of the notebook, most of which we have here, are from your work with that group on this project.

Townes:

Yes.

Nebeker:

And I was interested to notice that in a couple of these things that you did you have them witnessed by a couple of people. That occurs first on page 7. Maybe we should look at the pages as they come and see what you recall of this work.

Townes:

I saw a lot of Les Peterson, and Black. I'm afraid I don't recall a great deal about this. They were trying to study oscillators and potentialities of the microwaves. I was working with them, so if I got an idea, I tried to work things out. If I got an idea, I'd obviously write it down.

Nebeker:

It looks like you were taking a tube that Parker had experimented with.

Townes:

I was trying to analyze it.

Nebeker:

Your own analysis? Or did you also do some other testing?

Townes:

I must have helped do some experiments on it, too. I don't think I took the lead in experimentation this early. I was just watching other people and working with them.

Nebeker:

The effort in the electronics tube department was to get a microwave generator?

Townes:

Yeah. To get good generators. I'm afraid I don't really remember any details of this particular tube. Obviously I worked on it and obviously intentionally. Let's see, I tried to work out theory. I guess I was probably a little more theoretically inclined than some of them. But what I was trying to do was learn about it and understand it and work out theory.

Nebeker:

I think it's very interesting that, just a month or so after arriving at Bell Labs you had come up with an idea for a high-frequency generator. Do you remember if anymore work was done on this?

Townes:

Well, the only thing I clearly remember is that there was an oscillator which I kind of invented, which I thought looked promising. I did not try it out. I'd moved on to another department by then. But Jim Fisk, who was in the tube department, was asked to try it out. Jim tried it out and said it didn't work. So that was the end of that. [Chuckling] But that's the one that I remember. I think that probably went further than others and there may well have been a patent or something on it, I'm not sure.

Nebeker:

But that wasn't this particular one we're looking at here on page 7?

Townes:

I don't think so, no.

Nebeker:

Is this part of the Bell Labs practice, that when you get an idea that might lead to something that you write it up and get someone to witness it?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_2.mp3)


Yes. I was instructed about patents, you see. They're very patent conscious. So if you had an idea that you thought might be patentable, why, you witnessed it, just automatically. Anything that seemed like something really new and possibly patentable, why, you'd have it witnessed. If you took it more seriously, you might write up a memorandum and send it to the patent department. You'd send it around to other people and the patent department so that they could look at it and figure out whether they thought it was worthwhile. The patent department would decide whether they wanted to do something about it. During the Bell Labs period I had a moderate number of patents. A dozen or so. A number of them were connected with the war work. Since then I've patented practically nothing. I did patent the maser, and then I patented or rather Bell Labs patented the laser. In all of the years since then have I patented one other thing. [Chuckling] This was the generation of very high-frequency sound waves by Raman scattering. I patented that. I think that's all. At Bell Labs it was customary just to be sure that everything was reasonably covered.

Nebeker:

Yes. Right.

Townes:

They didn't want somebody from some competitive company stepping in and getting a patent and then blocking them from using what was developed there. They weren't so much interested in selling patents, as they were in protecting themselves to be sure they had everything covered.

Nebeker:

Looking at the following pages--I see that 9 has more analysis of the idea presented on page 7.

Townes:

Energy exchange. Let me see now. One thing that may bear on the rapidity with which I was able to do some of this, is that I had done my thesis with W.R. Smythe who wrote a book on electromagnetism which is a standard text. He was a very tough taskmaster, and I took his course which is a kind of a course that winnowed out everybody who couldn't quite take it at Cal Tech. [Chuckling] He was proud that this would sort people out. A lot of it was problem-working. I ended up working all the problems in his new book which he was writing. So I had been very much steeped in electromagnetic theory.

Nebeker:

So this kind of analysis was relatively easy?

Townes:

This kind of an analysis came fairly naturally to me. That may be the reason that I could get going fairly fast here. I see here's another thing on Les Peterson.

Nebeker:

You're looking at page 13. All of this looks to be further work on the apparatus described on page 7.

Townes:

That may be. This may well be the thing which Jim Fisk then tried out in some modification. That first version of it doesn't seem familiar to me. This one does a little bit more. This talks about field gradients and so on. Okay, now then we come to the other department.

Nebeker:

I didn't copy the rest of the pages. So at least half of September and October and November you were working with Llewellyn's group.

Townes:

Right, yes.

Nebeker:

There's one other memorandum here that is from that period. I don't know if you were involved in this. It's the end of November, right about the end of your time with that group.

Townes:

No, I think I was not directly involved in that. I'm pretty sure I wrote up a memorandum about this kind of oscillator.

Nebeker:

This is the first of the TMs that the database had for you, dated January 2nd. Do you think that is this work that's in that TM?

Townes:

Yes. There we are. "Energy Relations of Electron Flow Through Nonuniform High-Frequency Fields." In the second part of this is the nonuniformity begins to come in. As I remember, the idea was that some electrons flow through nonuniform fields, and with that, there should be a transfer of energy from the electrons to the field. So I had worked that out and wrote a memorandum about it. Then Bell Labs had somebody try it out, and I'm not sure he was awfully enthusiastic about doing it anyhow, and it didn't seem to work and be awfully promising, so it was dropped.

Nebeker:

You were doing other things, and couldn't pursue it?

Townes:

I was doing other things myself.

Nebeker:

Do you recall if you were disappointed.

Townes:

Well, of course I was disappointed. But it wasn't a big thing. [Chuckling] I thought it was a nice idea and ought to be tried, and they claimed it didn't work. I'm sure it would have worked in some sense; but if it didn't work, you know, more or less right away, then that probably meant that it wasn't all that useful.

Nebeker:

Was it specifically for radar that that group was trying to get a high-frequency generator?

Townes:

Well, I believe that Bell Labs even at that time was thinking about transmission, and primarily about transmission of telephone and television signals.

Nebeker:

Right, we saw that one memorandum was about transmission.

Townes:

Yes. They had been working in this field for some time.

Nebeker:

As early as '37 Bell Labs was working on radar for the Navy.

Townes:

I wouldn't have known just when they started. But radar was not the primary push.

Nebeker:

Okay. It was just for communications?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_3.mp3)


Their normal technical work involved looking at new forms of transmission. Down at Holmdel they had a group that was working largely with microwaves. There was a large amount of microwave work down at Holmdel at that time, in transmission and tubes. Southworth was becoming pretty well known for microwave work, transmission tubes, various modes and so on. He had also detected some radiation from the sun. He had taken microwave radiation from the sun. So Holmdel was a big center for fieldwork and experimental work and large-scale systems. They laid out waveguides, and some liked them and tried them out; that sort of thing. So Bell Labs was looking forward to transmission at that time, and that was what this department was primarily about. It may well be they thought some about radar, but that was not the primary push.

Nebeker:

And that wasn't what you were thinking about when you worked on this?

Townes:

No, that was not primarily what I was thinking about. The radar work was probably mostly secret, and I wasn't in on the secret part. There just was no point in my being cleared for secret work since I was doing this work, which might be related but I wasn't directly involved or thinking about radar particularly. I knew Bell Labs was working on magnetrons. The British had brought over magnetrons, and it was basically their efforts that certainly improved the potential of magnetrons.

Nebeker:

I think, though, that was the following summer.

Townes:

Oh, maybe it was. Okay. You probably know the history a little more precisely than I do at this point. They brought over magnetrons, and Bell Labs started working on that right away.

Nebeker:

How was Llewellyn as a supervisor in that period?

Townes:

Oh, very good. Excellent. He was a very high-quality person and stimulating and very nice to me. I liked Llewellyn. I enjoyed that whole group.

Nebeker:

Did you have any regrets about moving on to another group at the end of that period?

Townes:

No. I was in an exploratory mode. [Chuckling] Ready to try anything, and see what the rest of Bell Labs was like. But I liked it there. But then after I left, well, I said, "Okay, so I drop this now, and I start doing something else." It was clearly an exploratory period, and I enjoyed that, seeing what different things were being done.

Magnetic material research

Nebeker:

We know that the 7th of December you were working on this other case. You moved to the next group, and that I don't have identified very well.

Townes:

I think that was a tube group. Now actually it turned out I worked only with three groups before I got this reassignment to do radar bombing work. I worked with the magnetics group under Bozorth. And I worked in the tube group.

Nebeker:

Was that under Davisson?

Townes:

I'm trying to remember. Mervin Kelly had been head of that but then had moved on up into research director or something like that at that point. Let's see. I can visualize the person who was head of that. It was not under the famous Davisson, no. No, he was in physics. The electron tube group was not in physics again.

Nebeker:

Well, in the April of 1940 Bell Labs phone book, you're listed as being in Department 1180 with Davisson as the head of that group.

Townes:

That was primarily a nominal thing.

Nebeker:

So you were in a smaller group within that, do you think?

Townes:

Yes. I never worked with Davisson directly. I knew him well. But I worked with Bozorth who was head of the magnetic materials in the physics department--it was a solid-state group, but in the physics department--working on magnetics. Of course Bell Labs was very famous for magnetic work at that time. So that's where I was. Perhaps Bozorth reported to Davisson.

Nebeker:

We don't have an affiliation here for a long time. I mean, I assume that was the first three months you worked on your assignment. The next three months there's nothing there. Except we do know that it was this case which had the title "Fundamental Development of Electron tubes for Wire transmission Systems" that you were working on the next three months, December, January, February. Could that have been Bozorth's group?

Townes:

No. No, this was the development of magnetic materials.

Nebeker:

That was the third group you were working in?

Townes:

No, this was over in the old tube department.

Nebeker:

Maybe the name will appear in the notebook pages.

Townes:

Yeah, yeah, maybe. Let's see. My paper might mention something like that.

Nebeker:

You were talking about the theory of cathodes sputtering paper?

Townes:

Yes. Let's see my footnotes in the last part of the paper. Here's the name. Rockwood is the person I was working with then.

Nebeker:

Rockwood?

Townes:

He was not my boss, but I worked with him. As I say, I can visualize this guy, but I can't dig up his name. If you could find out who was head of the tube department at that time, I think I was officially reporting to him.

Nebeker:

All right.

Townes:

The vacuum tube department.

Nebeker:

I guess I don't have that department listed.

Townes:

I think that probably this fellow was over in electron tubes for wire transmission, though it was just basically electron tubes. Electron emission from filaments was a very important part of any electron tube in those days.

Nebeker:

Okay. Pages 21 to 46. Since there are so many gaps here, 21 starts right there. This was the first of your pages from that department, the electron tube.

Townes:

What I remember about this is Rockwood had made some measurements of sputtering which really then determined the lifetime of some vacuum tube in the long run. He had made measurements of lifetime and the function of various characteristics of tubes. After again looking around and seeing what people were doing and learning some things--and I may have written a few things about other things--my primary effort then was to try to understand his experimental data. So I did a theoretical analysis. See, in three months' time it's pretty hard to get an experiment going and doing anything real. I was moderately skillful at theoretical analysis. So the way to learn things seemed to be best to go to work on the theoretical aspects of things. Rockwood probably didn't have a Ph.D. degree. He was just a good engineer who was doing experiments and finding out things. But he had not done any analysis really, and this looked like an opportunity. He had some good data, and it seemed to me an opportunity to try to understand it.

Nebeker:

Is this the beginning of the work that led to the electron sputtering paper?

Townes:

Yes, that's right. That's what I'm talking about.

Nebeker:

And that goes for a number of pages here.

Townes:

It had to do both with the lifetime of filaments and the lifetime of these gas-discharge tubes.

Nebeker:

Do you remember if this was regarded as useful to that group? I mean, the fact that you wrote it up as a TM and then an article suggests it was.

Townes:

Yes, I think they appreciated it. Yeah, I think they appreciated it as the beginning of some understanding of what was going on. Hence that would allow them to engineer systems better, and to plan systems. I'm not sure really that it led to a lot of practical applications, but at least it gives some understanding of why the parameters worked out the way they did and what things one had to look for. So, yes, I think they were quite appreciative.

Nebeker:

Do you recall your feelings at the time? Although in a sense you're doing physics with both these first two groups, in a sense you're doing engineering. In your first case the main work was designing some generator, and here it's analyzing some experimental data. Do you recall your feelings at the time? Did you really take to this work, or did you feel that, oh, I want to be doing physics?

Townes:

Well, I would say I enjoyed this kind of work. It's the type of engineering which is so close to physics. It's not development of a big system or something like that, which I later did. Rather it involves trying to understand things. It's quite basic engineering, fundamental engineering, I would say. How do the electrons and fields interact, and how do surfaces interact with electrons? And so on. So that was close enough to physics that while I really didn't want to stay in those departments permanently, I felt it was quite interesting. I had no qualms about looking at that nor any lack of interest in seeing what was going on. But that was different from, building a system to put in an airplane. That's a very different kind of engineering, systems engineering.

Nebeker:

What do you recall of the second group, the electron tube group? Were you located on West Street?

Townes:

That's right. It was a little building right across from the main building on West Street. There's a little cross street there, and it was just over the other side of the street. It was a rather low building, a modest structure. It had been taken over by the Bell Labs from some other industry--I think I recall it being referred to as the old biscuit factory. It was just a rather large open building in which a lot of tube experimentation was going on.

Nebeker:

Do you recall whether there was a particular urgency for a certain type of tube or certain type of improvement that that group was interested in?

Townes:

I think at that point with respect to radar. The urgency was there, and people were thinking about that and what could be done. But otherwise it was a question of getting longer lives and higher gains and more current density. Also one of the big pushes--although now it's gone completely out of the window with transistors--was to get gas-discharge tubes which could operate on 24 volts. You couldn't get a gas-discharge breakdown in the argon-neon combinations at that low a voltage. But they would like them to operate on 24 volts because that was a battery voltage, and the voltage that many telephones use, and they wanted to use them as switches in remote places. I was even asked specifically to look at that at some point after I'd been in this tube department for a while. I think Harvey Fletcher asked me would I specifically look at that.

Nebeker:

He was Director of Physical Research?

Townes:

That's right, yes. But this was after I left. I left the other things and then was to work with Wooldridge for a while. And that was my assignment, to see if I could figure out a way of producing tubes that would break down at low voltages.

Nebeker:

This is later, not the third of the three?

Townes:

Yes, this was later. So that was a very specific thing I was asked about. Generally, that was just in the normal course of their needs. In the normal course of their needs, they needed tubes of all kinds. But in addition to that, the radar was beginning to be quite an issue, and people were concerned about it and figuring out what could they do, and how could they do it. I was not myself involved, though.

Nebeker:

Yeah. Of course this was in the so-called Sitzkrieg part of the war where nothing was happening in Europe. You've said earlier that you had this practice of changing your room every three months. Was that three-month period chosen to coincide with these three-month stints in different departments? [Laughter]

Townes:

I don't know. There might have been some unconscious connection.

Nebeker:

This was when you moved up near Columbia University?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_4.mp3)


That's right. I characteristically like variety. I like change and variety. It's a lot more interesting to see something new rather than stick with something that's old. New York's an interesting place, and I debated where to live in New York. The first place obviously was Greenwich Village. That's near Bell Labs and an interesting place, and so I lived there for three months. I didn't want to just get acquainted with Greenwich Village and nothing else. So that's when I said, "Well, okay, I'll move every three months. I'll just plan to move every three months and try another place." I was a young bachelor, and there was no problem about it. I just picked up my stuff and left.

Nebeker:

They didn't require year-long leases in those days?

Townes:

No. I was renting a room by the week basically. [Laughter] I just put all my stuff in a small trunk and got in a taxi and moved to another place. I moved up near Columbia as the next stop after that. Then I moved into Midtown, and I guess about that time, then I got caught up in the war. Then I was shuffling back and forth. I got married in '41, and we lived in a number of places then as I moved back and forth from Florida, back up north again. You'd have a room or house generally. Frequently I would give up my rented quarters and then come back to a new place because I was down in Florida long enough. I could save money that way.

Nebeker:

Yes. How long were those periods in Florida? Do you recall?

Townes:

Well, they were sort of a couple of months at a time.

Nebeker:

That long?

Townes:

Something like that.

Nebeker:

Okay. So is there any more that you want to say about that second group that you worked with?

Townes:

No, I don't think so. I think maybe I've covered that.

Nebeker:

Then it must have been in March, in '40 that you moved to a third group because the notebook records this as a different case. It's development of magnetic materials.

Townes:

Magnetic materials. Yes.

Nebeker:

And this is Bozorth?

Townes:

Bozorth, right. Now Bozorth was specifically a physicist, and I worked fairly closely with him. He headed the group. I don't think I did any particularly creative thing there. I was mainly learning. [Change to Side B of Tape] The magnetics department was quite a sophisticated department since Bell Labs was pretty ahead of the game in magnetic materials at that point. I learned a good deal from it. I was not thrilled by it because it seemed to me a little confining and a little limited in terms of scope. But on the other hand, it was interesting so far as it went, and a high-quality department.

Nebeker:

I only have one page, it looks like, from this case called "Development of Magnetic Materials," page 48 there. This is it. When I didn't copy pages it was because there were mainly just numbers there. There you see bridge measurements on inductance. Looks like you were taking measurements yourself?

Townes:

Yes, I was taking measurements. I may have had some idea about laminations and tried to do something there. I remember a man named Cioffi. Cioffi was working in that group at the time. Then there was still present a fairly elderly gentleman who had first discovered this very highly, very easily, magnetized material. I forget what it's called. [Chuckling] This was Bell Labs's great breakthrough. Anyhow, he was still around, and he was a grand old man, and he just sat there smoking cigars, I remember. Well, he'd made a great breakthrough, and that was it.

The effect of war on Bell Labs, 1940

Nebeker:

I know once the war started the pace of work at Bell Labs was intense with a lot of overtime and short vacations and so on. What do you recall of this period?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_5.mp3)


Well, it was a very pleasant period for me. For one thing, I had some money. I was paid $3,016 a year; I remember it very well. It seemed like a lot of money to me. As a student I had lived on about one fifth of that. Certainly I had some money, and I was in New York, and I was making a lot of new friends, and we'd all go out to lunch in Greenwich Village. They knew all the good restaurants. I was exploring New York. At some point I was taking music lessons up at Juilliard, too. So, while I worked hard, it was not something that I had to do at nights and Saturdays. I could work more or less at what would be a normal pace. It was a very pleasant time. I made a lot of new friends, and met people.

Bill Shockley was there, and Dean Wooldridge and Jim Fisk and a lot of other good friends that I've known for a long time. Plus a number of these engineers that I enjoyed very much. Bozorth was very kind to me, and Llewellyn. Les Peterson. I got to know a different group of people. Given my exploratory instincts, I found that very interesting. They were really quite sharp people and they also knew something about political things. I remember one engineer who interested me particularly. I believe he had a Czech background. He had a brother who visited in the Soviet Union and heard some of the Stalin trials, of that period, and swore that they were real, that these people would confess that sure enough they'd been betraying their country and so on. It was convincing and so this was a great plus for the Communist cause, you see, that these people admitted they had been doing terrible things against the government. His brother had been there and witnessed it and it was real. Along with the Nazi crowd this was one of the things that was being discussed. How real was this, and what was Communism, and what wasn't.

Nebeker:

Of course it was also in April of 1940 that the Blitzkrieg overwhelmed the Low Countries, and France was forced out of the war. Do you remember that month?

Townes:

I remember the invasion, but I don't place it in time or exactly what I was doing during that time.

Nebeker:

I was just wondering if you recall your feelings. It must have been a surprise to people how quickly France fell, and it seemed that England was almost knocked out of the war.

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_6.mp3)


That's when people really began to get very worried that we were going to have to be involved somehow. There was talk about it all along. But nevertheless that of course made it much more severe. I know the Bell Labs people were talking about what could they do to help besides getting ready and trying to help the British. They were looking for places where they could pitch in. They were very, very good in acoustics because Harvey Fletcher was an acoustics man. Then somebody under him was also very prominent in acoustics. They'd gone to the Navy and said, "We're experts in acoustics. We know that you're very much involved in the submarine problem, and can you think of anything that we can do to help?" This was before we got into the war and the Navy told them, no. We've got very good people, and everything's under control. Thank you very much, but we think everything's taken care of. Then some ships began to be sunk and they were called in.

Nebeker:

It may also be that the attempts to put technology to work in World War I hadn't been, except in a few cases, as great a success as they were in World War II. So the military services weren't as willing to accept.

Townes:

I think that's right. I think it was the chemists who came through in World War I. Not so much physics and engineering.

Electron physics research

Nebeker:

Then there's sort of a mystery case here that you worked on. Its number is 35887. It's the next page there beginning in mid-April of '40. You worked for quite some time on this case. That number isn't in the Bell Labs database now. Maybe you can tell what this work is by looking at some of these pages. On page 62 you're calculating the probability of ion capture and an ion capturing an electron approaching a metallic surface. That sounds like electron tubes.

Townes:

Yes, I think I know what this is, but let me see. Let me just check through this a little bit more.

Nebeker:

It doesn't look like this is still magnetics work.

Townes:

No, no, it's not. I think I know what it is. I'm just reviewing it.

Nebeker:

Pages 57 to 120 in this notebook are apparently part of Case 35887.

Townes:

Yes. Okay. My memory comes back to me now. I had not realized that I had spent as much time on this. I remember working on this only rather a short time, but apparently it was a much longer time than I realized. I know I left Bozorth's group to start working with Wooldridge. This was work with him. Wooldridge was then doing work in electron physics. He worked on secondary emission, for example, from surfaces. So I was assigned to work with him. This was part of the thing that I mentioned to you, that I should look into the possibility of getting gas-discharge tubes which would break down at much lower voltages.

Nebeker:

At 24 volts?

Townes:

Right. So then I began studying fundamental mechanisms in discharges. I had not recognized at all that it lasted so long. [Chuckling] My recollection is that it was a relatively short period.

Nebeker:

Right. Page 121 is the first with this new case number [37641]. Just the last few pages of this notebook. But it looks like similar work, doesn't it?

Townes:

Yes, it does. I think it's closely related work. And why it's a new case number, I don't know.

Nebeker:

So this was still from, it looks like, mid-April until the end of the year anyway. If we include the work on case 37641 this makes eleven months for this work.

Townes:

Now that surprises me, though.

Nebeker:

I mean, you may have been working on something concurrently, although there's not a notebook to show that.

Townes:

Yes, now let me see. Well, you see here is the 5th of March 1940 [37063-2]. That's magnetics. Here is May 31st.

Nebeker:

Right. Unfortunately there were two pages in between there that I didn't copy, but the first date on the new case number [35887] was the 17th of April.

Townes:

Oh, 17th of April. Okay. That's what I thought.

Nebeker:

And then you're working on this case until the end of November [35887], and then there don't seem to be entries for December of '40. And then the early part of '41 you're working on this closely related case, evidently.

Townes:

Let me see when that paper was submitted. I got the idea for that paper from Rockwood's results when I was with his department. But the paper was written after I was working with Wooldridge.

Nebeker:

Could it be that's what you were doing in December of '40?

Townes:

That may well be.

Nebeker:

Because the technical memorandum was dated January of '41.

Townes:

Yes. Well, I see I finally published it. But I sent it in in '44 so that it was a long delay before I sent it in.

Nebeker:

But the TM is dated January of '41.

Townes:

Yes. I remember Wooldridge was my boss at the time that I was writing it. The technical memorandum was just signed by me, was it?

Nebeker:

I haven't seen a copy of it, but it is in your name alone.

Townes:

It's funny the things you remember. But what I remember is that I was going to have Rockwood on the memorandum, or at least on the paper, and I had initially done that. And Wooldridge said, "Well, now look, Rockwood hasn't really done anything about this. You're being unfair to yourself," and so on. [Chuckling] I said, "Well, you know, I used his data." "No, you shouldn't put his name on it." I remember that, and obviously I was reporting to Wooldridge at the time. I remember a few things here and there, but I had no idea I'd worked on this as long as I did.

Nebeker:

Was this work that is on these pages here related to that paper, that TM?

Townes:

Yes and no. The primary object of this work was to try to understand the breakdown voltage in tubes. You can see discussions and experiments and so on of looking at ways of maybe studying breakdown and maybe making breakdown at lower voltage. That was the primary object. However, I was writing the paper in part during this time. When did the memorandum come out?

Nebeker:

It's dated 6th of January '41.

Townes:

Sixth of January '41. I was writing up a paper for publication using that as a base at some point while I was with Wooldridge. The paper was actually submitted quite a while later. Let me see if I can reconstruct this. My assignment with Bozorth may well have been cut short a little bit. My fourth assignment was to work with Wooldridge, which was again directly in the physics department. I was asked if I would look into the possibility of getting this low voltage. How soon that was after I started working with Wooldridge, I don't know, but probably fairly soon. So I was working in that field most of this time until then this defense problem came along.

Nebeker:

Right. That starts a new notebook which is dated 28 February of '41. First entry was the 3rd of March.

Townes:

That event I remember well. I think for one thing I never made very much progress in this discharge tube business. Obviously I was doing some work and looking at things, but I never had any great success, and maybe that's one of the reasons I don't remember it so well. [Chuckling]

Nebeker:

How was that group, and how was Wooldridge to work for?

Bell Labs seminars on physics research

Townes:

Oh, Wooldridge was excellent. He is a very intelligent person and sensible. He was doing good work himself. He gave me pretty complete freedom, and I would tell him what I was doing every once in a while. I remember interactions with Wooldridge very pleasantly. It was not a close interaction. He was just my supervisor. He was highly thought of in the Lab.


Audio File
MP3 Audio
(143_-_townes_-_clip_7.mp3)


I can't pinpoint the starting time but somewhere along in there Mervin Kelly had another idea which was very unusual for an industrial firm at that time. That was to have a seminar so that the people in the Lab could keep up with some of the latest things going on and continue to get educated. He also started what became known as Kelly College, which was a way of letting the technicians and others in the Lab who'd come in without degrees get a degree which was also a very useful function. The seminar was, I think, quite unheard of in industry at that time. He picked out a certain number of people--there were about eight or ten of us--who were to meet together once a week and basically have an afternoon off to talk about some aspect of physics or related matters. We could do anything we wanted to. We could read scientific papers, we could invite somebody to come and talk, we could read through important new books or something like this. Whatever we wanted. And the Laboratory provided tea and cookies, [Chuckling] which was again quite unheard of at that time.

Nebeker:

Was this Kelly's idea that the latest results in physics had relevance to the engineering?

Townes:

Yes, and that his people ought to be kept up to date. He had brought in--solid-state physicists, too, feeling that, well, solid-state certainly had something to do with the Bell Laboratories' business. I think part of this time when I was also working on vacuum tubes, I also looked at sputtering on switches a little bit.

Nebeker:

I notice sputtering is listed a couple of times in these pages.

Townes:

I looked at the wear of relay switches a little bit. I don't think I made any headway on it. But that was one solid-state problem obviously. [Chuckling] Electroconductivity through wires was another one. Transistors was another one which was to come along later and was not then in view. Certainly thermistors was another one. Thermistors were known at that time, as they had been invented by a chap there at Bell Labs. So Kelly as quite aware that solids might really make a difference to the business. He felt, that solid-state physics was a new field and he ought to hire some solid-state physicists, which he did. He hired Shockley who came in just a few years before I did. Foster Nix was another solid-state physicist there who was in our seminar group. And Dean Wooldridge, Jim Fisk and Walter Brattain were part of the seminar. That was a very stimulating group within the Lab.

Nebeker:

Was this physics seminar especially directed to solid-state physics?

Townes:

No, no. It was just anything we wanted to discuss. Anything we wanted to take up. We generally took up things that were close to our own work or things that might be close to our work, so they were frequently related. But they were just supposed to be the latest things that we found interesting and could take some time off to work on. That was going on sometime along in here and continued into the time after the war. It was so successful and worthwhile that similar things were being done even after the war. I remember we eventually studied Pauling's book on chemical bonds.

Nebeker:

It was throughout this period a group of a dozen or so?

Townes:

Yes. About eight to twelve people. Something like that. A lot of the most active research people. But Kelly had picked them out himself as to who was going to be involved. [Laughter]

Nebeker:

You felt privileged there, I imagine.

Townes:

I felt privileged to be in that group, and it was a privilege. A very good group of people.

Nebeker:

I'm puzzled by this long period with Wooldridge's group. It looks as if you did your little apprenticeship in different departments and then settled into that group.

Townes:

Well, I think that's right. I think maybe this was sort of supposed to be my fourth group, but it was a group that I was going to be more or less permanently attached to. I found it quite congenial, and I felt, yes, sure, that's a very reasonable thing for me to do. This applied physics to understand ionization and surface interaction of electrons and things of this type.

Nebeker:

That was congenial to you, that kind of work?

Townes:

Yes, I found that fairly congenial.

Nebeker:

You imagined that you'd be continuing that kind of work?

Townes:

So far as I knew, I would continue that for a while. So I went about it in a fairly extensive way. I had not remembered that it had lasted this long actually. I guess that's because there were no crises or anything; I just rolled along. [Chuckling] I think also I didn't have any great success. But then it was kind of a longer-term thing. These other things I just dipped into and I wanted to do something, so I did some analyses. This I was approaching on a kind of a longer-term basis.

U.S. decisions about involvement in World War II

Nebeker:

Do you recall the political-military situation in this period? There was something of a stand-off with Hitler unable to invade England. Of course this is before the invasion of Russia in June of '41. But I think throughout this period the United States was becoming more and more involved with Lend-Lease. Do you recall this period?

Townes:

Yes, I do in general. There was a lot of discussion about what was going to happen and should we do this or should we do that? And what Roosevelt was doing. Roosevelt obviously was trying very hard to help the British, and one could argue that he was trying to find ways in which we could get involved gracefully because he felt we had to be involved. But a number of people opposed that.

Nebeker:

What was your feeling at the time?

Townes:

I felt it was likely, but still kind of unthinkable. Kind of unthinkable that we should be in the war.

Nebeker:

Did you think the United States should do all it could to stay out of the war?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_8.mp3)


I thought we ought to do as much as we could reasonably to help but not overtly enter the war unless something else happened. That was my feeling. If England had been attacked at that point in a way which was completely threatening to it, I don't know what I would have said. It probably would have been too late for us to get going even. [Chuckling] I was not exactly opposed to doing something overt. On the other hand, I was not highly in favor of it either. I didn't see a way for us to get involved in any very clean way. Even though we might help them in minor ways, we might help out in some particular attack or something, but how could we reasonably get involved? In a sense it was very lucky that the Japanese attacked the way they did, and that swung the whole nation around. Because I think a lot of people were troubled by the same thing: Why should we get involved in Europe's troubles again? Is there a real reason? Well, yes, there's a kind of a reason, but there's no occasion. What could you back it up with?

Nebeker:

Well, it could have been the same sort of thing as happened in World War I. That is, more German U-boat attacks with escalating tension.

Secret research on computers' use for anti-aircraft guns

Townes:

There might have been, and people were thinking of that, and those kinds of things were happening. The Germans were being pretty aggressive, and those kinds of things were happening, and they were talking about what could get us in again. But mostly I think everybody wanted to help but just didn't feel there was a good reason for actually jumping in. There was, again, an increasing interest in the technology. I think by then Lovell's group must have started, and that was highly secret.

Nebeker:

Whose group?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_9.mp3)


Lovell's. He had the idea of doing anti-aircraft aiming using analog computers, electronic computers. The whole thing started by his somehow getting interested in--and maybe he'd done something on this before--using potentiometers for computing. Basically you rotate a potentiometer at a given place, and you get a given voltage, and so you vary voltages around in an electromechanical computing system. It's completely analog.

That led him to suggest using this for anti-aircraft guns. Anti-aircraft, that was a very hot topic at that time because the Germans were bombing Great Britain. So he was immediately gotten started on that. So he felt he had an idea and wanted to do it. It was a new idea. Everything else was mechanical and optical, and here he's going to put in some electronics and some electronic computing, as primitive as it was, and it seemed like an advance. So he made this anti-aircraft system, and it worked. Now I didn't know a lot about it.

Nebeker:

Where was that work being done?

Townes:

That was at Bell Labs.

Nebeker:

West Street?

Townes:

Bell Labs, West Street. I realized that was going on, I knew some of the people who were doing it, and so I knew sort of in general what they were doing. But it was secret, and I didn't have much to do with it. The computing business, of course, was very different then. George Stibitz was a mathematician; you may know about him. I remember very well his rolling around a relay rack. This was after I was over in New Jersey at Bell Labs; it was during the war. He was rolling around a relay rack to show people how you could compute with relays. He said, "Look, you've got to do it digitally. Otherwise you'll never get any precision. You have to do it digitally." Well, he was doing it with relays. [Chuckling] A rack full of mechanical things. [Laughter]

Nebeker:

Doing simple addition.

Townes:

People were not highly impressed with that. [Laughter] They weren't highly impressed, but, in principle you certainly can get more accuracy. We were doing a similar thing with analog computers, basically potentiometers. Lovell had the idea of shaping a potentiometer so that for a given angle you could get a more arbitrary function. You wound the potentiometer on a card of varying height so that the resistance varied, not linearly but with some other kind of functional form.

Nebeker:

Right. So it's something like a function generator?

Townes:

That's right. So that was a start.

Interaction between engineering and physics research at Bell Labs

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_10.mp3)


Now one other thing I would have to say about Bell Labs: I think it had a very good atmosphere in not differentiating between physics and engineering in any strong clear-cut way. They had a physics department, and they had various kinds of engineering departments. They had a chemistry department. But there was a lot of interaction. Engineers would work in the physics department a bit, and vice versa. There was a good deal transferred back and forth, and a good deal of interplay. There was not a sharp distinction between the two. I think that was a very good atmosphere from that point. Lovell, for example, was an acoustics physicist. Then he started doing this, and Bell Labs then assigned various engineers to help him out. They made a practice of hiring physicists and then transferring them into engineering. During that period engineering schools did not teach a lot of fundamental physics. Cal Tech was one of the few, and Cal Tech did it simply because it didn't have much of an engineering faculty. [Laughter] I think that's the reason they did it. I was a student there, and the engineering students would take a lot of fundamental physics because they had a small engineering faculty, so the engineers at Cal Tech learned a lot of fundamental physics.

Nebeker:

Well, also the 'thirties, exactly the period you were at Cal Tech, was a period of real change in EE curricula across the country. Until the 'thirties it was fairly rare even for the electrical engineers to get Maxwell's theory in any comprehensive way. But then in the 'thirties at MIT and Cornell and other places, a lot of physics came into the EE curriculum. So one can certainly understand why if in the 'thirties Bell Labs wanted sophisticated fundamental engineering, they would go to the physicists.

Townes:

Yes. Right. They'd convert the physicists. They'd bring them in, and they'd put them in an engineering department normally. I was a rare exception. I never figured out why I was so privileged, but generally they would just put them in a department wherever they wanted them. I knew a lot of my friends who went on into engineering and management picked it up at Bell Labs. The old-time engineers weren't very much trained in fundamental physics. Fundamental physics was coming in more and more. Now they get a lot of it, of course.

Nebeker:

Right.

Townes:

There was a strong interaction within Bell Labs, and lots were transferred back and forth. There was just no sharp division.

Nebeker:

So when you were working on some of these problems with electron tubes, there were people around you could ask if you had some question in electronics engineering?

Townes:

Oh, yes. There were engineers who'd been making tubes in a practical way for the system. There were also some physicists mixed with them. So it was a fairly effective kind of mixture.

Radar bombing system research

Nebeker:

I've read about how you were quite suddenly told--it must have been in March of '41--you were going to be working on the radar bombing system. You started a new notebook; the first entry is dated the third of March. You start work on Case 37616, which again wasn't in the database. From what we can see here on these pages, it's a servomechanism probably for a radar system. What do you recall? This is the next notebook.Townes:

Well, what I remember is I guess Harvey Fletcher, Wooldridge, one other, and myself were playing the lead. We were to report to Kelly. He had a new assignment [for us]. Although Wooldridge may have been tipped off a bit about it, I had known nothing about it at all. So I think maybe Kelly came into the office at that point, and he said, "Well, this is your new assignment. The war is unfortunately likely to hit us, is approaching, and we've got to help out. I want you to start working on a system for bombing by radar, using the new techniques which Lovell had been working on for anti-aircraft guns." Lovell and other people felt that this could be useful and workable. We would design a system using radar which would then allow bombardiers to use this system rather than the classical Norden bombsight.

Nebeker:

Was this all of Wooldridge's group that was suddenly reassigned?

Townes:

No, I'm not sure how big a group he had at that point. It wasn't a very big group. I can't remember if there was anybody else directly from that group. But there might have been.Nebeker:

Who did you report to then in the beginning in this?

Townes:

I reported to Wooldridge. Now whom did Wooldridge report to? I'm not sure whom Wooldridge reported to in the beginning. Later he reported to Walter McNair. But I think that was a little later in the war. At that point he may possibly have continued to report to Fletcher. I don't remember. Anyhow, we were called in, and next Monday we were starting our work. [Chuckling] Well, that was pretty sudden and unexpected, and I wasn't accustomed to being treated that way. I talked with Wooldridge. He obviously had been tipped off a bit about it, but it had been somewhat sudden to him, too. We then tried to learn what Lovell was doing and how he was doing it. We were also assigned some engineers to help us out. For example, I knew very little about amplifiers at that time. I don't think Wooldridge knew anything about them either. There was a good practical engineer who said, "Well, you can consult me whenever questions like this come up." Part of the time at least he sat in the same room with me, and he would help us out. He was really working for Lovell, as I remember, but he would be available anytime we needed him. Then they assigned a mechanic. At one point Sid Darlington worked with us. I don't remember whether he came in at the very beginning. I kind of think he did. Sid Darlington was a mathematician, an applied mathematician. Again, the mathematics department was very accustomed to doing engineering-type analysis. They worked a lot on noise and acoustics and things of this type. Sid Darlington was youngish then and a very capable mathematician. So he would help us out. Now our goal was to try to build something within one year's time, and you could never do that these days. But we were supposed to build something, and obviously everything was rush.

Nebeker:

How many people were in your group?

Townes:

I think there were about six, including technicians. I remember a technician, an engineer, and a machinist. And I think Sid Darlington, myself, Wooldridge. There may have been one or two more people, but that was it. So we just first sat down to design a system. Wooldridge was very good at it, and he was a very bright person. We all worked together. Sid Darlington--I think he was with us that early-- Joe Burton was another person who worked with us later; he was a physicist converted into this. I think he came along somewhat later.So we sat down to try to learn the business and design a system and begin construction. It was based on the technology that Lovell had worked out, which was relatively simple. He explained the main ideas of electromechanical computation, and you had to decide what equations you wanted to solve and what to do and how to solve them. The radar part of it, though, was not specifically ours. That would be done out at Whippany where there was a radar group. We were to collaborate with the people at Whippany, who were developing the radar.

Nebeker:

You were still at West Street?

Townes:

We were at West Street. They were developing the radar and already had things underway. They were making radars for radar purposes. Whippany was a big center then for the radar systems. Newhouse was the person, an engineer, in charge of that. Frank Goss was an engineer under him who worked on this. So the group, if you include the radar group, was larger. There were probably another half a dozen people there. They were already somewhat along in developing radar, not specifically for this but doing things that were easily applicable to this. What we had to do was to design an overall system. Newhouse did not report to Wooldridge nor vice versa. It was a parallel team, but we worked closely together. So we went out to Whippany to try to learn things out there, see just what the radar would do and what it wouldn't do, and then design a system. We built it, and we put it together. We had technicians to help us out in building circuits. [Chuckling] It was my first real experience with amplifiers, so I learned a lot. Learned a good deal about electronics and systems. The first system we built I'm not clear just when we got it ready--but we set it in a plane and got it ready sometime during the winter or spring. I remember it was cold, and we were flying out of Tampa, Florida.

Nebeker:

Could that have already been the winter of '41-'42? That would be less than a year after you started.

Townes:

Well, yes. It was the winter of '41-'42, or the spring of '42. I was married by then. I'd gotten married in the spring of '41. Since all this was secret, I wasn't supposed to talk to my wife or anybody about it, and some people were fairly careful about secrecy then. But she went down to Florida with me, as did Wooldridge's wife, and we spent some time there on the beach. I remember it being cold and rainy. [Chuckling] I'm sure it was winter. At this point I couldn't pick out the exact date.


Audio File
MP3 Audio
(143_-_townes_-_clip_11.mp3)


We did our first bombing exercise, bombing a ship that was anchored offshore. One of the curious things is that the Norden bombsight was so highly thought of at that time--and that made it so secret--that even though we were supposed to be designing a bombing system they would not tell us anything about the Norden bombsight. We couldn't see it, we couldn't know what it did. We asked our Air Force representatives "What kind of precision do you get? What kind of precision do you need?" They would just say, "Just do the best you can." People talk about the Norden bombsight dropping bombs in pickle barrels, and they had great stories about it. But the actual accuracy was not all that good, and particularly in Europe where the poor bombardiers had to make a long straight run in order to drop the bomb. We knew that that was a danger, and part of our effort was to try to see that one could maneuver and compute at the same time. In order to get the position of the plane relative to the target, you had to know the height and the wind speed, and of course the ballistics of the missile, and the direction of the plane, and its velocity. So basically you needed to calculate the wind speed and the drift from observing the ground. We argued right from the beginning we should be able to maneuver and do that. As long as you kept track of the plane's speed and orientation, that you could calculate the wind speed without that, and it would not make the poor bombardier run a straight course. So part of the idea was to do that, give him more flexibility, and part of it was to be able to bomb at night and through clouds when he couldn't see anything.

Nebeker:

Right.

Townes:

The very first run we had, we had a colonel who was a very nice person; we had a lot of different pilots during that period. Four of them in particular during the war, and three of them were killed in accidents. Fortunately, we never had an accident. But three out of four of our pilots were killed in accidents, and they were generally test-pilot types. This colonel flew us, and the first bomb we dropped we had a run at an altitude of 5,000 feet. 5,000 feet was a reasonably high altitude at that point. We dropped the bomb, and I quickly dashed up to the cabin to see what happened. It missed by about a hundred feet. And this colonel, who wouldn't tell us anything about the Norden bombsight--yes, sure, he'd used it. He couldn't say a thing about it. So then he said, "That's a damned good shot, if you ask me." [Chuckling] That boosted our morale a great deal, and gave us our first real information on the accuracy then.

Nebeker:

Why was it necessary for you to accompany the equipment on the test flights?

Townes:

This was a prototype system, and absolutely nobody else knew how to run it.

Nebeker:

You couldn't train some bombardier to use this prototype?

Townes:

We probably could have, but in any prototype system, there are always going to be problems. Something always breaks down.

Nebeker:

Somebody would have to be there.

Townes:

You'd really have to know the system. So in checking it out the engineers who'd designed the system really had to be there. This was not a production unit. We never made production units. Our job was to make a prototype system, and demonstrate it, and check it out, and see that it worked. Then it could be duplicated. Then somebody could be trying to use it. We could have trained somebody to do it, but we had to be there anyhow, and it was much more effective for us to operate it and see just exactly how it was going. So Wooldridge and I were sitting back watching the radar signals and guiding the plane.

Nebeker:

Were one of you acting as bombardier?

Townes:

Yes, except we had an automatic release. Basically we were watching the radar tuning the equipment, and the instruments, and watching everything. The needle showed the pilot how to turn. The pilot had a needle, which allowed the pilot to move it back and forth as much as he wanted, and then at a certain time he had to turn on course. Put that needle exactly in the middle. That was his job, and then it automatically released when the computer told it to release.

Nebeker:

Yes. I see.

Townes:

We were watching the instruments primarily and the radar signals. We tracked on radar. That was another thing. You had to track on the radar signals. There was not an automatic tracking. You had to track on the radar signals. That's how it worked. This is where I learned about servosystems.

Nebeker:

This is the very first page on that project. On the second page we can already see some of the mathematics involved.

Townes:

Yes. That's right. Some of the mathematics for the equations which the system solved.

Nebeker:

So you're learning at this time a lot about systems and electronics and also this quite new field of analog computation.

Townes:

Also stability of feedback circuits. Hendrik Bode was at Bell Labs at that time, and he was a member of the mathematics department. I had to think about and learn Bode's theories and get well acquainted with built-in feedback systems. Well, yeah, this is a set of equations.

Nebeker:

On page 6 there's a heading: "Maximum Velocity Necessary for Motors"?

Townes:

These are the motors that drive the potentiometers, I guess. Let me see.

Nebeker:

It says, "Potentiometer must move complete circle in about 20 seconds." In the middle of that page.

Townes:

Yes. Right. We had the motor geared down so it was set to calculating the speed and distance.

Nebeker:

Now this is real system engineering work.

Townes:

That's right. It's invention as you go along, too.

Nebeker:

Sure.

Townes:

But it's system engineering, and it had a function to fulfill. So that's what we were doing.

Nebeker:

Did you find this interesting to find ways for this kind of function?

Townes:

Well, yes. It was interesting. It was new to me, and so interesting. Sure. I felt in itself it was interesting. I was pleased to learn about it even though this wasn't what I wanted to do in the long run. I was engaged with the analysis of it. I adapted to it, I guess, reasonably well. I didn't find it uninteresting.

Nebeker:

There is also some mechanical engineering, for example, on page 29 the calculation of the torque of the motors.

Townes:

Oh, yes. Right.

Nebeker:

On page 31, just two pages further. You've got a condition for non-binding of worm gears that you've got to check. I'm sure your Cal Tech training in mechanics proved useful in all of this. [Laughter]

Townes:

I'd taken a mechanics course but no engineering. But these are things which are fun to work out. And I was forced to think about them.

Nebeker:

Page 33, exhibit possible slowing of the worm gear.

Townes:

Yes. Various torques. Boy, I don't remember all of this.

Nebeker:

That seems fairly sophisticated mechanical engineering if you look at page 33 there. That tooth of the gear and the forces that are involved there.

Townes:

Well, you know, it illustrates that principles apply really very broadly, scientific principles. [Chuckling] Or engineering principles. Apply very broadly across a lot of different fields. If you know something very thoroughly, you can make some headway in a lot of different areas. I had had fundamental physics and mechanics and electricity and magnetism.

Nebeker:

What can you tell me about page 40?

Townes:

Oh, yes. Here's another one of those schematics.

Nebeker:

Its head is: "Consideration of Current Drawn by Various Potentiometers and Accuracy Necessary and Corrections." These are the calculating potentiometers?

Townes:

Yes, these were the calculating potentiometers, and there's the necessary fractional accuracy.

Nebeker:

Was your group actually building these potentiometers?

Townes:

Yes, we wound them.

Nebeker:

And then you'd help in the testing of them?

Townes:

Yes, yes. I think the Lovell group was already doing some of this. We just learned from them. I don't remember interacting with the Lovell group very strongly after the first few weeks on this kind of thing. But they were always there, and they'd been in the game for, oh, probably a year and a half or two before we started. They had already made a successful system.

Nebeker:

I'm trying to understand how this analog computer worked. Looking again at page 40, are you developing a set of calculating elements with these different potentiometers which you would then use as necessary for the particular calculation?

Townes:

Yes. Right. If you look at the next page, it says, "Computing Circuit." SD. That's Sid Darlington. So he was with us from the beginning.

Nebeker:

That's right, because that's March of '41.

Townes:

Yes. Sid Darlington. Now Sid was very familiar with electronic circuits. He'd been very close to Hendrik Bode and the people in the mathematics department. But again, they were really doing engineering. It was a mathematical kind of engineering, and Bode's Theorem, came out of that. Sid Darlington was very familiar with those, which I wasn't initially. So he was helpful there. He evidently drew this circuit or had it drawn up; probably he did it himself after we had laid it out and talked about it. Basically you end up with a steering circuit.

Nebeker:

What is being steered?

Townes:

What is being steered is the airplane.

Nebeker:

Okay.

Townes:

This is the needle. You end up with a needle that tells the plane where or how to turn. Particularly just before the drop, where to turn. That depends on the crosswinds and various sorts of things where the plane is and the ballistics of the bomb. There is also a release circuit.

Nebeker:

I see.

Townes:

When these things come together and go through zero, that releases the bomb. You put in voltages representing the different variables, and you come out with a voltage that tells you how to turn the plane, and a voltage that tells you when to release the bomb. That's a computing system.

Nebeker:

These input voltages. Where are they coming from?

Townes:

We had an altimeter, which is basically a pressure measurement, in the plane. We had a air velocity measuring device.

Nebeker:

I don't know the instrumentation technology at that time. Could you get a voltage out of an altimeter immediately, or did you have to work that out, too?

Townes:

At that time we set it by hand, the altitude. The plane kept a constant altitude.

Nebeker:

I see.

Townes:

You kept a constant altitude, and you set that by hand. Initially I think the velocity was constant. Then later we mechanized it so we'd have a varying velocity. I think initially the velocity was constant. So you would read that and set it in by hand.

Nebeker:

Okay. Just turning a dial.

Townes:

Yes, yes. The plane kept the constant altitude, and constant air velocity, but could turn. Then you'd put in the various constants of the bomb by hand. You'd track the radar signal electrically anyhow on a scope, you'd put cross hairs on the scope.

Nebeker:

The bombardier, that's his job?

Townes:

He tracks the target, and to track a target, you turn things, and that puts in voltages. By tracking a target, that gives you then the information from which you can calculate the wind speed and direction and all of that.

Nebeker:

I see.

Townes:

This is the initial system here.

Nebeker:

That was arrived at very quickly.

Townes:

Yes.

Nebeker:

Was the group going before you joined?

Townes:

No.

Nebeker:

So in the space of a month you had this whole thing worked out.

Townes:

In one sense this is a simple system by comparison with what people do these days. It was new then, but nevertheless a simple system by comparison to what we do now. As the war went on, we put in more and more complications. We put in systems for both optical and radar guiding, for example. The last systems were built for all of those characteristics. We put in more complications as time went on, perfections and so on. But this was the first system, and that's the one we tried out first down in Florida. We moved fast. It was a small group and a good group of people. We just had to decide things, and so we did.

Nebeker:

On the next page, there's a servo for rotating the antenna about a horizontal axis. Is this for automatic tracking?

Townes:

Yes, this was to keep the antenna on the target as the plane rotated and turned. As the plane advanced. The target was not always directly in front of it. I think the antenna probably had a potentiometer on it, too. I've forgotten details. The axes for three degrees of freedom were not terribly accurate. The beam width is moderately wide. Then we had relays for controlling some of the power. Mechanical relays always give us trouble. [Chuckling] Mechanical relays and vacuum tubes, and those were the best that were available at the time. And by the end of the war I felt we had as many as we could afford. Otherwise, the troubles with them would be too frequent.

The system was about as complex as you could afford to build at that time, with the tubes frequently going out. We had to have very controlled qualities for the tubes. The tubes had to be linear and accurate; this was a d.c. computation. So the d.c. drifts from the tubes had to be small. We even had a special tube with a special number made with those specifications.

Nebeker:

That could be done with a tube group at Bell Labs?

Townes:

I've forgotten who was making those tubes at that time. I think RCA perhaps. But whoever was making those tubes, it was not Bell Labs.

Nebeker:

I see. Some manufacturer. You said, "We need these characteristics."

Townes:

Some manufacturer. We simply put on the specifications and asked them if they could give us tubes with that specification. And they said, "Well, yes, okay. That'll be X more expensive, and we'll have to make a special class of tubes." They gave it a separate tube number for that tube. They were then available for other uses, too. But it was much more carefully controlled.

Nebeker:

That's interesting, that the existing tubes didn't meet that need anyway.

Townes:

Yes. Well, we were pushing on everything for precision, and that was one of them.

Nebeker:

This is mid April of '41. There's a system sketched out again.

Townes:

Yes, I see Sid was the one who sketched that out again.

Nebeker:

And on the next page, "Measured by J.H. Kronmeyer".

Townes:

I don't remember who Kronmeyer was.

Nebeker:

Maybe a technician with your group?

Townes:

I don't think so. I think we may have gotten somebody else from a different group to make some measurements for us or something. I don't quite remember his name. This is a distance measure, we're checking out a potentiometer.

Nebeker:

Now this is in the summer of '41. Now let's see. I was looking at page 100. That looks like figuring different ways of calculating something. I guess we can see the dates here on the next page, two pages further on; it's 118. Yeah, this one. "Bombs dropped."

Townes:

Oh, yes.Nebeker:

We've got some dates here, February of '42.

Townes:

Okay. I knew it was in the wintertime. I guess that's when the time was.

Nebeker:

Now of course that's a couple of months past Pearl Harbor, and that must have given added urgency to getting the system working.

Townes:

Yes. Now you see altitudes set. We set an altitude. We read the actual altitude. The pilot didn't always stay right on altitude.

Nebeker:

I see.

Townes:

We had a set-in altitude for what the actual altitude was.

Nebeker:

I see. It indicated air speed and true air speed.

Townes:

Yes. One of the standard calculations was for true air speed, and then, from tracking, ground speed.

Nebeker:

So you had some air speed indicator, and then the system is calculating air speed. Is that right?

Townes:

Well, the air speed indicator depends on air density in part and temperature so you had to correct it. I think this is just standard correction for the way the meter read.

Nebeker:

Oh, I see.

Townes:

There was no way of measuring ground speed except by doing the radar tracking, which gave a measurement. I think this here is just the standard correction to the air speed meter.

Nebeker:

Okay. And you can see that there's a column, "Range Error."

Townes:

Yes.

Nebeker:

And I don't know what that next column heading means.

Townes:

That is the left-right error. Left-right deviation error.

Nebeker:

Oh, deviation error maybe.

Townes:

But that's the distance left-right. And you see the very first one. [Chuckling] Hundred-foot range error and zero deviation. The next one wasn't as good.

Nebeker:

But then you had some very good ones.

Townes:

We would drop a bomb, and then we would come back and take another run and make it from a different direction.

Nebeker:

Are these averages then? There were six runs on February 12th.

Townes:

That's right. Six runs on February 12th.

Nebeker:

February 27th, there are five runs, I guess. And so on. I see.

Townes:

That's right. So those were our bombing runs.

Nebeker:

It looks impressive, If you look at these range errors down this column. You must have been quite pleased with the system.

Townes:

Well, we didn't know what was supposed to be good. [Chuckling]

Nebeker:

I would imagine that in a project like this, even to get it working approximately would be a great achievement.

Townes:

Yes. Oh, sure. Well, we were thankful it was working. [Chuckling] We'd been working hard. We didn't know what was supposed to be good and what was supposed to be bad. We just did as much as we could in the kind of time available, and got it as accurate as we could with the kind of techniques we had.

Somewhere in here, at some point, I think it was probably later than this, I actually did a statistical analysis of these errors in the system. My impression is that it was one of the very first times that anybody had tried to add up a lot of minor errors in a system to predict the overall performance of a system, just on a probability basis. I made measurements and estimates of all the errors and all the different kinds of parameters and elements. Then added it all up. I remember joking at the time, well, this is what the mathematics says it should be. [Chuckling] Of course nobody knows whether this kind of thing works out or not, but this is what the mathematics says it ought to be. Well, it turned out to work out quite well. That kind of systems analysis is common nowadays.

Nebeker:

But they weren't then?

Townes:

Nobody that I know else had done then. They all seemed to ignore it. Nor did they want to trust it very far. [Chuckling]

Nebeker:

It sounds like something that Sid Darlington would take to.

Townes:

Yes. Well, Sid, I think, was ready to believe it.

Nebeker:

The idea was that you have these estimates of all these individual errors for the different components of the system.

Townes:

Yes. Then you add the sum of squares, assuming Gaussian errors, which you don't know to be Gaussian, but nonetheless it is a first approximation.

Nebeker:

Then see what error distribution you have for the system as a whole?

Townes:

That's right. That was the idea. Get the system error from the component errors. And it's now, very common. Everybody accepts it, everybody does it nowadays, if they're interested. But I think that was one of the first. I don't know of anybody else who was doing it at that time. But that was probably a year or two later than this.

Torpedo bombing

Nebeker:

Or page 130 or another page or two there's a geometry of torpedo bombing. Was that a special version of this computer for torpedo bombing?

Townes:

I don't remember.

Nebeker:

It may have just set the dials differently.

Townes:

I really don't remember. I must have been asked to look into that, look into the possibility of torpedo bombing.

Nebeker:

But that was not the main purpose of this?

Townes:

No, no, we never did any actual torpedo bombing.

Nebeker:

I see.

Townes:

As I say, I must have been asked to look into it.


Nebeker:

A couple of pages further on page 140 you've got a heading "Second Derivation of Torpedo Bombing Equation." So you're working out what the calculator in the computer would have to do for torpedo bombing.

Townes:

Yes, yes. Right. We must have been asked to look at that. Maybe we just thought it could be useful and started doing it. I don't remember anything about it, so I guess nothing ever happened.

Nebeker:

Well, it went a little bit further. On the next page you've actually sketched the circuit for torpedo bombing.

Townes:

That's the system. I see.

Nebeker:

And you've signed it and dated it.

Townes:

I see. [Chuckling]

Nebeker:

So it looks like it's your contribution.

Townes:

Must have been.

Nebeker:

You don't recall its being built?

Townes:

No, I don't recall its being built.

Nebeker:

There's a second torpedo bombing circuit. So these are the last pages of that notebook that I copied. There's an air speed meter. You've got some notes on behavior of plane in rough weather. Pitch and roll.

Townes:

Yes. Right. Must have correction with thermistor circuit.

Bell Labs work environment

Nebeker:

You said that you didn't have much contact after the first period with Lovell's group.

Townes:

No. Not so much.

Nebeker:

There must have been other groups there working on these kinds of analog computers.

Townes:

No. Well, I don't remember. I don't remember other groups. We continued to work all through the war, and we were in close contact with the Whippany gang. We worked quite closely with them. Lovell continued, as I remember, but it turned out we didn't need to have any special contact with him about other things. The potentiometers were going along reasonably. They were never terribly accurate or terribly flexible.

Nebeker:

They were doing the job.

Townes:

We used them as they were.

Nebeker:

What about the work environment? As I said, I've read about how short vacations were, something like an average of two days a year at Bell Labs during the war. And that it was common to work overtime and Saturdays. What about in your group?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_12.mp3)


We worked quite hard. I can't remember the vacation situation at that time. I probably took some vacation. I don't know. I'd have to think about it. I just don't remember at the moment anything. But we certainly were working pretty intensively.

Down in Florida where we were bombing and checking out the equipment, I was down there a moderate amount, coming and going. It was sort of a vacation in a way. Not always, but some of the time, my wife came down, and we had a child by then, a baby, and we would drop down and go on the beach. The plane was not always flying. Obviously the plane had to be taken care of from time to time, or something would break down. So we had a moderate amount of free time just from all of these days when we couldn't work. Some of the time I did some physics. Some of the time I did a lot of underwater swimming and diving, which was not common at that time at all. I got a kind of a faceplate so I could see underwater. There was no scuba diving at that time and no snorkeling. The result was that lobsters were very common just offshore, and I could go out and get a lobster anytime I wanted to. Lots of interesting fish.I wouldn't say that the work was all that oppressive. It was very intensive at times, and when the plane was working, we would do a lot of flying. Then repair the equipment in between times and fly again. So it was intensive at times, and yet there were also periods during that testing when you couldn't work. So we'd knock off, and I could do some other things. We would do something. I like natural history very much, and so does my wife. We'd go take a hike and look at birds or swim. I don't remember taking vacations, but very likely I took at least some vacations. And I can't tell you how hard I worked excepting that [Chuckling] I know I worked pretty hard.

Reaction to the atomic bomb and other military developments

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_13.mp3)


One thing I remember is that when the first atomic bomb was dropped, I was working. It was about midday Saturday, and I was working in a little hut in Whippany checking out some radar. I've forgotten what. But I remember this little hut, and I had a radio, and I heard that this bomb had been dropped. And it was an unknown kind of bomb that did a very powerful job. Well, I knew precisely what it was. A number of friends I had who had been working on the system were indiscreet enough to keep me posted on what was happening. I remember very well. And I said, "Well, I don't have to keep working today." So I shut down and went home. [Chuckling] So I was working on Saturday that day at least. I think it was fairly common. Even Saturday nights, and we were working hard.

Nebeker:

How closely did you follow the war itself?

Townes:

Quite closely.


Audio File
MP3 Audio
(143_-_townes_-_clip_14.mp3)


We, were directly concerned about the bombing effects and what they were doing and how successful or unsuccessful they were. Now I must say, none of our systems actually ever got used in the war. That was a disappointment to me, but to indicate to you how ready people were to help out--and I certainly was--after I had done this a couple of times, developed systems and they said, Yes, that's very nice. Now would you develop this kind of system? They always mainly wanted to go toward, shorter wavelength radars. I decided that maybe it would be more useful for me to just drop out and go over and be in the Eastern Theater. I was thinking I'd look into the possibility of going over and joining [General Joseph W.] Stilwell in China as some kind of technical assistant there. There was not a clear way where I could be a lot of help. But, I thought, maybe I ought to do that instead of working at Bell Labs and trying to build these things because they're not getting any use. My bosses then got wind of this, and they worked on me [Chuckling] very hard to stay. I didn't quite see the right kind of an opening there, so I said, Well, okay. I'll just keep on doing this.

Nebeker:

Had you been in communications at Bell Labs, it would have been easier to take such a position overseas.

Townes:

Yes. Perhaps so. If I'd tried hard I could have gotten a position alright. There were a lot of different factors involved, but I considered that quite seriously. I thought, maybe I really ought to be doing that rather than developing equipment at that point in the war. It was fairly late in the war. I was also kept on after the war. They wanted me to continue because we had systems in development which clearly were going to be used for airplanes, and they wanted them finished up. And Bell Laboratories was very eager for me to stay in that business. So I stayed on about six months after the war, and then they let me out on the basis that I'd found somebody to replace me, which they did. Wooldridge had been transferred immediately back into physics right after the war. But they'd left me in charge of doing the systems planning. Actually I should make this plain, too. That I think that after the first system, we were not in charge of the overall systems building planning in principle. There was a mechanical engineer, for example, at least in the latter part of the war, named Mottram, who was in charge.

Manufacturing, testing, and reliability

Nebeker:

What's the name?

Townes:

Mottram, M-O-T-T-R-A-M. He was a Bell Labs mechanical engineer, and he was in charge of seeing that the system was made and could be put into production. The Whippany people were in charge of the radar. We were in charge of designing and making a system that would work. Mottram didn't really understand the system very thoroughly. He was more of a production engineer. But we reported to him, and he was reasonable enough. We really did the design, but he just kind of looked after it to be sure that it would be manufacturable. He was in charge of the overall system, including the radar from that point of view. Now this first system, I guess we were completely in charge. But it was a prototype system. As we got closer to something that was likely go into production, then Bell Labs decided they ought to put a production engineer in charge.

Nebeker:

Can we look at the pages I've photocopied from the next notebook dated 28 October '42? Are these goals or achievements?

Townes:

"Items considered while at G.V." Now I don't know what G.V. is.

Nebeker:

Maybe these are goals.

Townes:

I think those are probably goals.

Nebeker:

And then on the following page, further goals are added. The first 20 are things that you were working on.

Townes:

Yes, these are a list of different items and goals. Now here I mention the accuracy to be expected.

Nebeker:

Yes, I was just looking at that.

Townes:

BTO. That means "bombing through overcast." Probable errors computed from memo. I wrote a memo, I guess. This sounds like a more preliminary thing and calculations I'd have made.

Nebeker:

What is this on page 19?

Townes:

"Change of resistors to be expected during potting.... Data taken by Fay."

Nebeker:

What is potting?

Townes:

I think that's probably getting the resistors set in a plastic which then hardened around them. I think that's what it was. This must be possible changes when you did that. Problems of changing resistance.

Nebeker:

I see. Now we're in February of '43.

Townes:

Yes, here's offset bombing. Well, that's a complication we put in, because not all targets would show up on the radar. You'd want to look at one target but bomb another one. That's the offset bombing. So that was put in as a further component. The system got more and more complicated as we refined it.

Nebeker:

And then on page 53: "Measurement Indicator for the Land Version of BTO." What's the difference between the over sea and over land versions?

Townes:

I'm afraid I don't remember. I mentioned Shank and Elmendorf. Bob Shank was involved with radar, and Elmendorf also. They were a couple of radar people. They had different groups of radar people to work with us, and they worked with us a number of times. This was just a way of making a kind of little more complicated setup. I don't know about the land version.

Nebeker:

A couple of pages further on, this one I was interested in. You've got things to check, and you have Newark. So there were some flights out of Newark?

Townes:

Yes, we flew out of Newark testing the systems. We couldn't drop bombs on the Newark area but we could fly.

Nebeker:

You were actually dropping sandbags or something?

Townes:

Well, they were real bombs in shape. In shape they were real bomb casings, but they were filled with sand to bring the weight up to just what the bomb would ordinarily be. So they were mechanically, so far as their flight through the air was concerned, exactly like a real bomb.

Nebeker:

You couldn't do that in the Newark area.

Townes:

No. [Chuckling] Go over the city and drop sand on them. So we flew out of Newark to do a preliminary check before we went on down to Boca Raton, Florida.

Nebeker:

Okay. Now that's July of '43.

Townes:

Now you see a pilot instruction: I tell him to hold the height within 10 feet, and the direction within a half degree.

Nebeker:

Here on pages 90 and 91 it's apparently some idea of yours that maybe was patentable since you have it witnessed.

Townes:

Yes, yes. I guess that's right. Patentable way of refining the instrumentation and so getting better readings.

Nebeker:

I just thought it was interesting that you were apparently doing some sensitive testing here in August of '42. With these different dates you've got on all these pages.

Townes:

Yes, yes. Precession rates, gyros. That's right.

Nebeker:

Do you recall this period?

Townes:

I remember Bell, the person mentioned there. We did a lot of testing in the laboratory of precision of the equipment and behavior and stability in the servos and all of that. We checked the system out quite well in the laboratory as a system. Then we installed it in the plane. There were plenty of things to be worked out.

Nebeker:

I was just noticing on page 94 that you have a test flight from 4:30 P.M. to 8:00 P.M. Three and a half hours. Were they typically that long?

Townes:

Oh, yes. Let's see, we were bombing something off Key Largo then.

Nebeker:

Yes, see August 17th, two flights total seven hours.

Townes:

Yes, the bombing runs were fairly long. I see I mention Goss and Soffel. Soffel was a technician. Warnock was an Air Force representative who was there to see how it was coming out.

Nebeker:

Was that unusual?

Townes:

No. Of course the pilots were Air Force test pilots, and they, in a sense, could tell the Air Force how we were coming. But they would send people from time to time to fly with us and see how it was coming out. That was not so unusual. One time we had a load of fairly high-level people--some Bell Labs people and some people from Washington--fly in our plane. The radar antenna housing got stuck. That is, you let the antenna down out of the body of the plane so the antenna could see out, and then you'd raise it back up where the antenna couldn't see in order to land. Well, it was let down, and we couldn't get it up. Everybody thought that all these bigwigs were going to have to jump. [Laughter] Oh, dear, what had we done now. All these older men, distinguished people, we were going to make them jump out of there.

Nebeker:

Couldn't you land the plane anyway?

Townes:

No, they said it was too dangerous. Too dangerous. So I managed to climb down in there with some wrenches and screwdrivers and get it fixed. It was generally motor-driven, you see, and it had gotten jammed, and they couldn't do a thing with it. [Chuckling] It was a heck of a job. Did it by hand and got it back up again. So we had visitors every once in a while. That was the most important group of people to ever visit us, though. [Laughter] Just the right time for something to happen.

Nebeker:

Well, we obviously can't go through all the different things you were doing, working on the system, but it helps me to get an idea of this work.

Townes:

Well, there were very extensive tests, as you can see.

Nebeker:

Do you remember the system as being fairly reliable?

Townes:

It was fairly reliable. Our primary difficulty was the tubes and the relays, especially the tubes.

Nebeker:

Would you test them separately?

Townes:

Oh, we'd check them on the ground. I checked them on the ground beforehand and individually checked all the tubes.

Nebeker:

I see.

Townes:

And then replace whatever ones were out of spec.

Nebeker:

Before every run you'd test all the tubes?

Townes:

We'd test the critical ones. I'd test the critical ones. It was sort of a zero test game check which you can do rather quickly. I would say the principal breakdowns generally were vacuum tubes, there we so many of them in the system.

Nebeker:

How many would you guess were in this system?

Townes:

I think we had something like seventy or a hundred tubes. It was my feeling that that was about the limit at the time.

Nebeker:

There were also relays you said that gave some problems.

Townes:

There were relay problems. A relay sometimes would get stuck. There weren't as many relays, electromechanical relays. I don't think we had so many problems. I think the potentiometers worked fairly reliably.

Nebeker:

You had at least some servomechanisms for this system.

Townes:

Yes. Servomechanisms are primarily a question of getting them tuned up to the best performance and so on so they're stable.

Nebeker:

What were they doing? We saw the one that was directing an antenna.

Townes:

Yes. I'm a little hazy about this. But I there were potentiometers where you would servo the potentiometer to a given voltage, and then there would be another potentiometer on the same shaft, which would then do the calculating for you. There were several potentiometers on a given shaft. It wasn't that the potentiometer just turned, as a result of some external input, to a calculated position where it had to be and the potentiometer went there. It was servoed. Then another potentiometer on the same shaft did the additional calculation there and read off the voltage.

Nebeker:

Was there feedback with that first one to turn it until you got a certain reading?

Townes:

Oh, yes. That's right. Negative feedback. Suppose you have a linear potentiometer. It would be given a voltage that it had to match, and it would turn to that voltage and sit there. Then another potentiometer, which was not linear, would then read off certain voltages which came from somewhere else, and did this calculation. But we had a number of servos. Antennas were some of them. Potentiometers were others. I think that was probably about it.

Nebeker:

One could imagine with a complex system like this that it could be very frustrating, that it would be difficult to get everything functioning.

Townes:

Yes, that's right. Well, it was complex for the time. But I knew it very thoroughly because I built most of it myself with few associates of course, but I was quite familiar with all of it. The servicing was not a severe problem, but it did take some attention.

Nebeker:

I don't know if there's more to say about all of these test runs. A new notebook starts there. "Averaging Circuit," it says.

Townes:

Averaging circuit? On what page?

Nebeker:

Page 108. I'm still at the back of the previous notebook. This is interesting. At the back of this notebook you have some of the results graphed.

Townes:

Oh, yes.

Nebeker:

There is also, on the page before that, pictures of this target.

Townes:

Yes, right. That was a target constructed there. Just a kind of reflector.

Nebeker:

Is that to make it a readily-seen radar target?

Townes:

That's right. It had a wire screen on it. And as a corner reflector it has a characteristic that if you send a beam into it, the beam gets reflected back in the same direction rather than scattered in all directions. So that makes the beam quite intense. If you had just a radar scatterer like, let's say building, then you'd have some directions moderately intense, others weaker.

Nebeker:

Yes, I see.

Townes:

In this case it's like a cat's eye. You shine a beam of light into a cat's eye, and the cat's eye will send the beam right back to you. So the cat's eyes glow. They look like they're shining in the dark. This does the same thing, back along the same direction. That's a principle very much used in optics these days.

Nebeker:

Okay. I just thought that was interesting to see that.

Townes:

I see I got a photograph of it. I haven't seen these things in years. I'm kind of glad because it's interesting to see them again.

Other radar bombing research groups

Nebeker:

I'm talking with Charles Townes in his office at Berkeley on the 15th of September 1992. This is Rik Nebeker. I've copied these from the relevant case files at the Bell Labs archives.

Townes:

These of course are internal documents. You see AN/APQ-7, a memorandum written by Bob Shank. Now Bob Shank and Chuck Elmendorf worked together on another system. They were located in the Graybar Building on Varick Street, as I was with Llewellyn for a while. They were also developing radar and a radar bombing system, the AN/APQ-7 I think was theirs.

Nebeker:

I see.

Townes:

But we consulted back and forth. That was a rather simpler system, and these memoranda indicate that they had planned to possibly put some of our computing systems on their radar. They were primarily radar people, who adapted the radar to bombing. Whereas we were primarily working in bombing, computed bombing, and navigation and computing systems, and putting that on the radars, you see. They refer to our AN/APQ-10 for other forms of the computers such as the AN/APQ-10 type or UBS, these could be added with minimum amount of modifications at some later time to this. The APQ-10 was the one we were working on. The UBS was the "universal bomb site," we called it, which was to be both optical and radar and we also developed that. These are things that we were doing, and they were thinking of maybe having us adapt it to put it on their system.

Nebeker:

This is a memorandum from the 19th of February '44.

Townes:

'Forty-four, that's right. This is fairly late, and this is from Bob Shank. Now here's another one, and you see the people involved. Now there's Newhouse who was at Whippany and a radar man and in charge of the radars generally which we used, at least initially in the game.

Nebeker:

Was this Model 2?

Townes:

Mod-2, yes, of the AN/APQ-7. That's those memoranda. That group was building still a different system but I'd say primarily it was a modest adaptation of a radar system that they had designed for bombing.

Nebeker:

We have some of these numbers. The early work you were doing was on apparently the DU150550 bomb site radar. Do you recall that number?

Townes:

I never used those numbers very much. They may be official, and I may have had them in my notebook, but these were the names we used, the APQ-10 and the UBS and so on.

Nebeker:

Now this was Case 24839. Yeah, this was the AN/APQ-7 case which is different from this.

Townes:

The Eagle radar as I remember was a radar to be put on the B-29 as I remember. I think the MIT group was involved in the Eagle radar also. That sort of thing is mentioned here. Eagle-type radar, it says here, I think.

Nebeker:

So your involvement in that was adapting some kind of a computing system to that radar?

Townes:

That was the idea. We never actually did it. We discussed it. These are memoranda about our discussions. We never actually did that. We discussed it back and forth and looked into it in some detail and made some plans for it. This right here is a kind of a conference for getting ready for manufacture. Here's Newhouse again, the radar man, and Mottram who was a mechanical engineer, that was really in charge of getting us into production.

Nebeker:

Now this is October of '42.

Townes:

Yeah, this is earlier. I didn't realize Mottram was involved quite that soon. It may be that he wasn't really doing this system at that time, but was in another system and was an interested party.

It was February of '42 that we made the first test flights with the first system.

Nebeker:

It was March of '41 that you started it, so just under a year to a working system.

Townes:

Yeah. That's right. This is October. We are onto another system by then.

Nebeker:

Okay. Is this a 3-centimeter system?

Townes:

Probably. Now we talk about the BTO, "bombing through overcast." That probably was a 3-centimeter system. You see these are just statements about the status of various units. To what extent were they ready to be made.

Nebeker:

Your main involvement on that was again the computing system?

Townes:

Well, we did the intellectual design of the whole system, marrying the radar and the computer. But we did not directly design the radar. We were the systems people. We designed in detail the computing part, and hooked them up, and made some specifications about the radar. So it was the overall system, but the radar was a kind of an important subset that was done by radar people, generally people at Whippany. This is another kind of internal memorandum about what's involved and what parameters are involved and capabilities. These are how to make these potentiometers and the specifications for the potentiometers, the shaping of potentiometers. By then they were pretty standard, and we knew a good deal about them.

Nebeker:

Do you know if the work that your group did on these potentiometers was used by other groups?

Townes:

I don't know. I don't know of any substantial use, beyond Lovell's group and our group, of this kind of computation. It was originated by Lovell, but we extended it substantially. We used it right on through the war, and the last system I worked on really was blocked out by MIT and then brought down to Bell Laboratories and developed. At the very end of the war and after the war, the last one I was working on--and that went in the B-52 eventually and was a standard piece of equipment--that had some of these techniques in it. I lost track then. I don't know what happened after that. And of course other kinds of computing began to come in.

Nebeker:

But there was a short period there beginning in the war and lasting until the early 'fifties when analog computing was a big field.

Townes:

Well, I can't tell you too much about the history, particularly after I left the field shortly after the war. You know we were buried in our own thing, doing our own things. Certainly there were both mechanical analogs and electrical analogs, and we had combinations of all of them. We were inventing new combinations: various ways of loading the potentiometers, shaping the potentiometers, loading the potentiometers, using them in combination, to get more and more complicated functions.

Nebeker:

Except for your contact with Lovell's group, you were pretty much working on your own on it?

Townes:

I believe so. I don't remember anybody else who was doing anything close enough to this and importantly enough for us to worry about.

Nebeker:

It wasn't a case of secrecy requirements making it unlikely that you'd learn about other work?

Townes:

No, I don't think so. I think in this field people were fairly open. The Norden bombsight I mentioned was closed for a while until we were asked to make a system which combined optics and radar, and then they opened up the Norden bombsight.

Nebeker:

One radar engineer I talked to complained about the fact that during World War II often the same work was being done at maybe a research laboratory and Rad Lab because all that work was classified.

Townes:

I did not see that aspect of it. The primary people we interacted with were at BTL and the MIT group. The MIT group I always felt was open though they were rivals; they were competitors of Bell Labs. But I felt that there was a good deal of trading back and forth of information. I never felt any secrecy problem there at all or any other hesitancy. Everybody was interested in the war, and while they were competing and everybody was trying to have the best ideas, nevertheless I didn't see any impediments to the trading of information for that reason. Now other companies, for example, let's say Raytheon, I'm sure did not share completely with Bell Labs for commercial reasons, and would deal in a more or less standard commercial way. We didn't see very much of the other commercial companies, excepting those from whom we were getting parts and supplies. The people who made the Norden bombsight we saw a good deal of. I remember seeing something of Minneapolis-Honeywell. They had some kind of navigation and bombing system. We went out to Minneapolis and looked at that and talked with them. I've forgotten just what happened there.

Nebeker:

Did you have contact with Western Electric? Did they do any of the work for things you were working on?

Townes:

Western Electric was the manufacturer for AT&T at that time. If these things had ever been manufactured, they would have been made by Western Electric.

Nebeker:

But you didn't have contact with the production engineers there?

Townes:

Well, there were a few Western Electric personnel who had come in and looked at what we were doing, and talked about how it might be manufactured and tried to size it up. So they interacted with the Bell Labs people, but they were not systems designers. They were basically a manufacturing group, but they were interested in the fact that they might have to manufacture this, and how would it be done? And they looked at it and tried to evaluate it, and talked with us, and somewhat influenced our design from time to time as to what they could make, what their procedures were.

Nebeker:

I see. Did they tell you, That's going to be impossible to produce in numbers, that kind of a potentiometer?

Townes:

I always put specifications on the systems and I remember one time when Walter McNair (Wooldridge's boss) came to me and said, "Well, Charlie, I really have to change and relax what you've done. The Western Electric people have said it would be very, very difficult. So I've opened up your specifications on that. I just had to do that." [Chuckling] No apology from this guy.

Nebeker:

That's pretty interesting.

Townes:

So there was that kind of case where the manufacturers just felt that while something was possible, it was too difficult.

Nebeker:

It's interesting to know that your group did get some input of that sort.

Townes:

Yes. We did. This was the latter part of the war, and they were to be the manufacturers of those things.

Nebeker:

What about the rivalry you referred to? Were there groups at Bell Labs that you felt you were competing with?

Townes:

I don't know. I don't think so. Take the Shank-Elmendorf group, for example, in some sense they were competitive. They were making another system. But I don't think they felt any difficulty about that. It was a simpler kind of a system, and they were radar people. They were good friends. We talked a lot about things.

Nebeker:

Including the accuracy of your system?

Townes:

Oh, yes. We talked.

Nebeker:

Did you feel you were competing on that score?

Townes:

No, no.


Audio File
MP3 Audio
(143_-_townes_-_clip_15.mp3)


They were not in the same ball park in terms of complication of the function and the accuracy. But they were doing a useful job, and I would say while we were doing somewhat parallel things, there was a healthy interaction. I don't remember feeling any particular rivalry. The rivalry that I saw was primarily between Bell Labs and the MIT group. The MIT group was a pretty powerful group, and it was big with lots of good people. They had built some very simple bombing systems which actually got into use. They really started the radar bombing business so far as I know. It was a very, very simple kind of a system. Basically just fly in a straight line and use the radar for guidance. They were doing that, and that was in use. Then we went in for rather more complicated things, more precise things and more maneuverability. They were primarily still doing simple radar and bombing systems work. We were more complex and not as close to the war operations.

During the last of the war we were suddenly asked: "Well, would you redesign and build a system which MIT Laboratory had proposed and put together a demonstration in the lab at least?" That was a little bit of a blow, [Chuckling] because they had suddenly tried to jump ahead of us and do something still more advanced than we were doing, and had sold it to the military, and the military had said, "Well, of course, if we're going to have it manufactured, that ought to be AT&T not Lincoln Laboratory." So they got us to redesign it for manufacture. It had to be pretty completely redesigned, but used some of their ideas. And they had some good ideas. That was the last system that I was involved with.

Nebeker:

Do you think you could identify that with what the case number was? You can see these different cases. This is the HAB X-band BTO, and this is the Eagle APQ. And you also worked on this, the AN/APQ-34 fairly late. I guess this was the last of these that you worked on.

Townes:

It had to be that one, then. It had to be that AN/APQ-34. It had to be that one that was the MIT one. That's the K-band.

Nebeker:

It was this radar plus this system.

Townes:

APQ-34, right. I think that must have been it. I think that must have been the one that originated at MIT.

Nebeker:

The AN/APQ-34.

Townes:

Yes. As I say, a sizeable group of people were put together to redesign that and check it out, and then get it to Western Electric in proper form for manufacture. I think that was it.

Uses of radar systems

Nebeker:

In the front of your notebook by this case number, you put GPI Model 1, Ground Position Indicator, Model 1.

Townes:

I'm a little hazy about this, but I think maybe the Ground Position Indicator was a sort of continuous calculation of coordinates as to where the plane was with respect to the ground. That's why it was called that. At the same time, this was a system which had moved us from the original wavelengths of 10 centimeters down to 1-1/4 centimeters. The 1-1/4 centimeters I was rather opposed to, and some of it I suppose was emotion. I was fed up with building something and having it changed by someone else.

Nebeker:

And you could see the war was coming to a close.

Townes:

That's right. And I thought, well, these things weren't really going to get into use. In addition to that, though, I recognized that this was a wavelength that could be absorbed by water.

Nebeker:

Van Vleck wrote a paper on it.

Townes:

Yes, I think I first woke up to that as a result of his paper I then read. In any case, he wrote a paper on it which was an informal memorandum really that was passed around. A very good memorandum that was passed around.

Nebeker:

Where was he at the time?

Townes:

He was at Harvard, and he was working on noise and picking signals out of noise and things of this type. There was a Harvard group that was working on signal to noise problems and radio signals. He was a theoretical physicist, and somehow he ran into this and wrote a memorandum saying, "Well, this might possibly give trouble." That may have been the first time that I woke up to the problem. I read that with considerable interest, and then worked on it and tried to extend it and figure out how bad would it be, [Chuckling] and decided it was likely to be fairly bad.

Nebeker:

Did you come up with estimates for the likely range of the radar signals?

Townes:

Yes, I did. I made estimates of what it was likely to be from basic physics, and I estimated the most likely amount of water in the air. I talked with people about it. The Bell Labs people said, "This is our assignment, and that's what we're supposed to do. And we'll have to talk to the Pentagon and Lincoln Laboratory." I talked with Rabi who was a big shot at the MIT Laboratory--I keep calling it Lincoln--the MIT Radiation Laboratory eventually became Lincoln. I talked with Rabi who was a big shot there, and Rabi just wouldn't listen. I talked with people down at the Pentagon a bit. They set it aside, and they didn't know how to answer it. I remember talking with a British officer who was stationed over here. And he said, "Well, jointly we've decided things. The United States had decided to go after the 1-1/4 centimeters. It's already decided, and, well, we just have to go ahead." I was still pretty young then, and people just didn't want to listen. [Chuckling]

Nebeker:

I know one of the anticipated uses of these systems was for mortar location from the trajectory of a mortar. That requires a high resolution and relatively short range. So maybe this system would be useful.

Townes:

Oh, well, I wouldn't say the K-band radar was useless. But for a bombing system it was.

Nebeker:

I see.

Townes:

For a bombing system they had to have some range. That wasn't very useful. They were also going to put it on ships and longer-range things and all kinds of things. Well, it was just the rush of the time. It was a new field. People didn't know enough about it.

Nebeker:

Maybe there wasn't enough confidence in one or two physicists' calculations of the range.

Townes:

That's right. It was all theoretical. Nobody had actually shown this; it was theoretical. But on the other hand, the theory was basically very sound. Anybody with knowledge about it couldn't really be in much doubt about it. It's sort of factor of two uncertainty, but not much more. The uncertainty is largely the broadening of the line, the shape of the line due to collisions in the air. How much would the sharpness of the line be decreased? So if it was broadened out, then it's not quite so severe. However, at that point Japan was becoming our main target and the air over the Pacific was known to be quite moist.

Nebeker:

But there were experimental results that supported this?

Spectroscopy and the K-band radar

Townes:

There were experimental results on this in related spectroscopy, related spectroscopic results. One could make a rough estimate just from the size of a molecule; you can make a rough estimate. Rabi, as I said, didn't seem to want to listen to the argument. Nevertheless, during the latter part of the war, the Columbia Radiation Lab where Rabi was very influential made a big experiment to actually measure the shape of the water line. They had a big room resonator, resonating at 1-1/4 centimeters in multiple resonances. They had wet air and dry. Rabi was the head of that laboratory so that obviously he did, in the long run, [Chuckling] decide he'd better check up. They did some of the first work then in absorption of microwaves by a molecular line. It wasn't the very first work. The first work had been done before the war on ammonia at Michigan just from the point of view of a physics experiment. At Columbia, they measured the shape of the water line and at atmospheric pressure. All of that was part of the picture of my waking up to the potentiality of a microwave spectroscopy.

Nebeker:

Is that the first time you did serious work in that area, calculating absorption?

Townes:

I had done some spectroscopy in my thesis at Cal Tech. It wasn't a primarily spectroscopic thesis, but I'd done some spectroscopy.

Nebeker:

That kind of spectroscopy?

Townes:

No, no. This was normal, visible spectroscopy. I'd taken a course in molecular spectroscopy at Cal Tech--essentially to get educated. But I was not in close contact with the field. But this got me interested, and as I started looking into it, I saw the potentialities more and more.

Nebeker:

Is it fair to say that if you hadn't worked on that K-band system that your interest in microwave spectroscopy might not have been awakened?

Townes:

I think that's quite right. That's what did it. I looked at it carefully enough that I saw the potentials, and otherwise I wouldn't have paid that much attention to it. I would have gone off in some other direction. Now maybe some other direction would have been fruitful, but on the other hand I certainly wouldn't have gone in that direction.


Audio File
MP3 Audio
(143_-_townes_-_clip_16.mp3)


No, that really started the whole thing. Of course the availability of K-band equipment was very important. It became very cheap, lots of it, and it had been developed pretty well. There were good local oscillators, klystrons, and waveguides and detectors. What really got me interested in microwave spectroscopy was looking carefully at the shapes of these lines and thinking about the theory. Since we didn't now much about the shapes of the lines, we had to look very hard at theoretical predictions. Van Vleck and Weisskopf had published a paper on spectral line shapes, and that was a very good basis for trying to understand them. But nobody had ever quite recognized that by going to lower pressure, the line just got narrower and narrower without getting weaker, and the center of the line stayed just as intense. The line simply got narrower. That was so surprising. In terms of any spectroscopic practice nobody had ever seen that. They'd pump gas down and found that the lines got weaker. That was the normal experience. That experience was primarily because the width of the lines was limited by the spectrometers, not by the molecules themselves. So the spectrometers have a certain width. As the line gets narrower, and you're looking at that full width, well, the total fraction of energy absorbed in that width keeps going down.

But if you had a better spectrometer, then right at the peak of the line it'd be just as intense. Well, nobody had ever experienced that, and they couldn't quite believe it. In fact, when I tried to sell the idea of working on this to Bell Labs, Jim Fisk, who was then head of the physics department, felt that it really couldn't be right. I argued with him about it, pointing out that's exactly what the theory said, and I don't think you can get around it. He then called on Arnold--oh, dear, I can't think of his name now. A theoretical physicist who had worked at Columbia and was familiar with their work on the water line. He'd transferred to Bell Labs, and Jim Fisk thought highly of him and trusted him. He said, Look, would you look into this thing and find out and tell me, can Charlie be right? And Arnold look at it and said, "Yeah, that seems right."

Nebeker:

Who was it who looked at it?

Townes:

[Chuckling] I'm trying to think of his name. It's a Scandinavian name, and his name is Arnold Nordsieck.

I primarily mention that because it was waking up to this whole new field which most other people didn't realize was there. In fact, they largely doubted what was happening. That's completely because I got interested in looking into the problem. It's completely because of that. Otherwise I would not have awoken to that either.

System improvements

Nebeker:

Maybe we could look at these. There are not that many pages from these last couple of notebooks. You probably have copies in that folder there. The Notebook 19729, if we could start with that one because we went through the other ones yesterday.

Townes:

Okay.

Nebeker:

This is the HAB X-band BTO radar [case 24822-1]. There's more of this error analysis. You seem to have done a fair amount of that.

Townes:

Yeah, well, that was important to us.

Nebeker:

I was also struck by some examples in these pages of your doing what today would be called numerical analysis. I don't see it right now, but I think that must have come up for a lot of physicists and engineers.

Townes:

I see in here some calculations I made for Bob Shank. I had business with Shank, and then I covered various things. He was worrying about some of these things.

Nebeker:

These were better approximations.

Townes:

I see here's a kind of additional arrangement of potentiometers that we thought would give the kinds of functions approximately that one needed. So I was trying to advise Shank on some of the things he was doing even though I wasn't working with him directly. There's a cosine amplifier, I see.

Nebeker:

Can you explain what that is?

Townes:

Yes. [Chuckling] Well, we had a hard time approximating this particular function, as I remember.

Nebeker:

The cosine function?

Townes:

Yes. And I don't remember just why. But somehow we couldn't build a potentiometer that approximated it sufficiently well. So what I did was to say, well, let's run the potentiometer along. Then when the voltage gets to a certain point, you put in a new resistor at a certain place which then modifies it and modifies the function. You keep putting in these additional resistors as you go along; then you can get any arbitrary function you want. Now how do you put in these new resistors? Well, as the voltage changes, you slip in a diode which cuts on or off. So this is an arrangement with a lot of different resistors and vacuum tube switches, which when you come to a certain voltage, this thing flips in, modifies the function, and then you can flip in something else. Obviously you go along, say, in time, and then you just flip in a batch of resistors, and you can get any arbitrary function you want.

Nebeker:

This is something like a hybrid between a digital and an analog computer?

Townes:

Yes, a bit like that. [Chuckling] That's right. Except that with the resistor then that carries out a certain slope in the curve. You're suddenly changing the slope in the curve, is what you're doing.

Nebeker:

I see. So you're approximating this curve.

Townes:

It's a series of linear slopes, sort of.

Nebeker:

I see.

Townes:

I remember some of my friends thinking I was a little bit crazy when I actually built it. They said, What? I said, it works. [Chuckling]

Nebeker:

Why did they think it was crazy?

Townes:

I think they thought that it was a little too complicated. Also, it was a new and additional complication for the potentiometers that they'd been working with.

Nebeker:

In other words, one potentiometer with the cosine.

Townes:

They doubted that I could get enough accuracy. By picking up just the right slopes and functions in a modest number of switches I did it, and they were easily impressed. [Chuckling] It really can be done.

Nebeker:

Do you recall if this was actually put into any system?

Townes:

I don't know. I'm not at all sure we actually used it in the long run. I remember very well one of my friends who was a real doubter, and he was persuaded. I remember also I had a very bad case of poison ivy at the time. [Chuckling] I see that's some more potentiometer arrangements.

Nebeker:

This is interesting. You've got some notes on APQ-10 tests. Do you think you took part in these tests?

Townes:

I essentially always did do the testing. I took part in any tests. I wasn't there 100 percent of the time. I think by then Joe Burton was much involved in the tests, and the engineer Goss was involved in the early tests, though he tended to fade out later. Joe Burton may well have done some of these tests, but I'm sure I was down there part of the time.

Nebeker:

On the next page you have all of these 17 adjustments to be made to the system.

Townes:

Adjustments and corrections to be made. Yes. I see that I've checked off some of them. Well, I'm still doing that kind of thing with our telescopes that I'm currently working with. [Chuckling] That's a list of all the things that need attention.

Nebeker:

Or the next page you've got "Items To Be Done" which is quite a list down at the bottom of that page.

Townes:

Oh, yes.

Nebeker:

This looks very typical sort of system engineering work.

Townes:

That's right. Again, this is not all that different from work I'm doing now, in a sense, on two telescopes used for interferometry in the infrared. They're much more complicated, and they use computers all over the place, of course. Nevertheless, they have checklists like this. [Chuckling] Lots of little things that need attention in the overall system. I remember that those were rather similar. I remember when I first proposed building a telescope because I had for some time been working in the laboratory just with rather simple, refined, but single elements and spectroscopy, and with lasers and so on. I had proposed this fairly complex telescope system to NSF. One of the reviews came back saying, "Well, can Townes build anything like that?" [Laughter] I guess he didn't know about my war work. There was a good deal of similarity. Any particular pages here that you want to discuss?

Nebeker:

I'm just flipping through this as well.

Townes:

All this brings back memories. I see the flight to Memory Rock and then Banana River on the coast. "AFC fair on approach but oscillating on search. S.B. Greenbow." You asked me if we had many troubles. Well, this is a kind of list of troubles. Now of course those were manageable, but we still had to pay a good deal of attention. Radar range was very poor, radar was not performing as well as it might at that time. On the other hand, it was rare that we went out and didn't have a reasonably successful bombing, but it was never as good as I would have liked. There were always things to be improved. But the system was fairly functional.

Nebeker:

Was it the case that you were trying to get to some predetermined accuracy, improving the system until you got there? Or was it more a matter of getting reliable service out of the system?

Townes:

It was a combination. We wanted as much accuracy as possible, but there was no cut-off point where it was good or it was bad. We tried to get as much accuracy as we could. We put in reasonable components and complications. In addition, we watched the serviceability of them. So it was a combination of the two.

D-Day

Nebeker:

The period we're looking at right now is April, May, June of '44. Of course D-Day was there on June 6th. Do you recall that day?

Townes:

I don't recall so sharply the end of the European phase of the war. What I do recall, as I mentioned yesterday, is the dropping of the first bomb on Japan because I felt that.

Nebeker:

I thought maybe the announcement of the D-Day invasion might have seemed important.

Townes:

Oh, I remember the invasion, of course. But then that lasted a while. No, I remember the invasion very well and the build-up to that. And there was a lot of talk about when it would happen.

Nebeker:

I know there was hope that Germany would be knocked out of the war before the end of '44, which of course didn't happen. I'm just thinking this might have affected your attitude.

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_17.mp3)


I don't think so particularly. We still had a war to win, and the Japanese were pretty fierce, and that was a big job. Nevertheless, after Europe fell, it was clear we could concentrate more on Japan, and it had to come to the right end at that point. I felt that there was no chance that we would lose. Well, I felt that toward the latter part of the German period, that we were getting ahead. It was a matter of time. There'd still be a lot of sacrifice. But after Europe fell, why then there clearly was no doubt about the outcome. Nevertheless, the Japanese were so fierce that it was hard to say how long it was going to last and how many lives would be involved. I don't remember any kind of let-up on the part of anybody in that period because of those things. Various people differed in their dedication. Generally the country was pretty solidly behind the whole affair. Some people were making money on it; others were just sacrificing. Overall I think there was pretty much of a unity in the sense of being willing to put up with things until the war was over.

Mechanical analog device

Nebeker:

I had a question about page 74. It's talking about a system for the bombardier to determine ground speed. And the second paragraph reads: "However, since bombardier has to figure ground speed, probably okay to make him figure drift and ground speed on ABC computer." What is that?

Townes:

I don't know what that is.

Nebeker:

It looks like you're proposing a kind of slide rule system.

Townes:

We had a drift-distinguishing plane. That may be constructed to give tilt.

Nebeker:

Is that 0.5o?

Townes:

It looks like 0.5o. Within half a degree perhaps?

Nebeker:

I was just interested in that here you are devising a mechanical analog device.

Townes:

Well, we were open-minded. Whatever worked. [Laughter] The optical people--the Norden site was completely optical-mechanical, generally relied more on mechanical things. It may be that I was trying to adapt it as quickly as possible. I had no aversion to mechanical systems. As I say, just whatever seemed to work was fine. We were willing to use any method possible. But I don't know what the ABC computer was. Might have been something like Automatic Bombing Control. I just don't know.

Probable error calculations

Nebeker:

Okay. The following page has one of your more elaborate figures. Looking at all those different components of the system, if you look in the right column, you're giving the error at 2,000 feet.

Townes:

Yes, I see. This is a pretty complete tabulation. Well now, then I see finally a total of 30 units which gives 275 feet probable error, 13.3 mills at 20,000 feet. I think this was maybe the calculation where I had added up everything. And so far as I know it's the first time anybody'd tried this. Just add up everything as the square root of the sum of the squares.

Nebeker:

Is that how you arrived at the final probable error?

Townes:

I think so. I looked at the estimated errors in each individual component.

Nebeker:

Is there a theory that when you have a series of errors that the final error is the square of the sum of the squares?

Townes:

Probabilistically. That's not perfect of course. That's making some assumptions that each one is Gaussian and independent which clearly it isn't. On the other hand, that's about as good as you can do. You really don't know the distribution of each one. So that's the kind of thing that's normally done these days with almost any errors, as long as they're independent. So long as they're independent errors, why, you could almost always use that method if they're Gaussian. That's generally not a bad approximation.

Nebeker:

Was that regarded as satisfactory, 275-foot probable error?

Townes:

Yes, 13 mills was considered good.

Nebeker:

Is that the angular measurement?

Townes:

That's the angular. That mill is 1/1,000 of the altitude. Mill is a milliradian. That was the standard language which the bombardiers used. The mill error tends to be larger as you get lower in altitude because some of the errors were fixed in dimension rather than angle. So 13 mills at that altitude there, that's quite good. Or considered good at the time. Now 20 mills at 5000 or 10,000 feet was considered generally as good as one could expect to get.


Audio File
MP3 Audio
(143_-_townes_-_clip_18.mp3)


By this time I knew the Norden bombsite quite well, and I knew it wasn't really all that good. It was a nice device for its time, but it wasn't terribly accurate. The biggest problem of all was that the poor pilots and bombardiers were sitting ducks up there if they operated in a time when they could see the ground and people would see them. They were required by the Norden to make a long straight run, and they were duck soup for anti-aircraft fire. The result was that they would try to come in as suddenly as they could, and they would probably drop their bombs before they should have, get out of the way, and try to dodge. The precision was probably much more affected by that behavior rather than the instrumentation itself.

Nebeker:

Or the next page you talk about various ways of coordinating the Norden sight with this Ground Position Indicator system. The idea, I take it, is that one can rely either on the optical, the Norden system, or a radar system, as conditions allow.

Townes:

That's right. Of course if you've got broad daylight, the Norden site could pick out a given object probably more precisely. Some objects don't give very good radar targets, and you have to do offset bombing which might be less accurate. So the optical pointing precision was clearly better when the optical conditions were right, and hence it was desirable to be able to use that. Now our system also allowed the plane to dodge around while guiding to stabilize the optics so that they would always point in the right direction. So we could continue to calculate and compute the wind speed as the pilot maneuvered around. Then at the last minute he could turn on. It was still dangerous to bomb in the daytime, but this was nevertheless a help to the bombardiers.

Nebeker:

Right. Do you know if that hybrid sort of system came into use during the war?

Townes:

No. It came along after the war.

Nebeker:

Okay.

Townes:

This was December '44 we were working on it. Circuits for computation.

Television

Nebeker:

Now this is from another notebook actually, this one here.

Townes:

Yes, I see. We jumped to '47.

Nebeker:

This is from your last notebook, and it's dated 2nd of August '47. Use of bombardment-induced conductivity for a television tube. This is apparently an idea you had for a new type of TV tube?

Townes:

Yes, I guess so.

Nebeker:

It goes on for three pages, and you've signed it and dated it at the end, although it doesn't seem to be witnessed. But it suggests that you might have thought of this as something that is worth looking into.

Townes:

Right. Well, I had forgotten all about this, frankly. It's vaguely coming back to me now.

Nebeker:

There was, black-and-white television at this time but not broadcasting. You describe here a new way to construct a picture tube.

Townes:

I guess I felt that I had an idea and maybe I ought to write it down. [Chuckling]

Nebeker:

What's interesting is that even in this microwave spectroscopy period after you've been at it for a year and a half anyway that you get an idea for a TV tube.

Townes:

I really don't recall now how that came up. I must have been reading and thought of some ideas that certainly struck me. I probably may well have had somebody talk to me about how difficult it was to make a good TV tube. [Note added later: I recall now my interest and that I obtained a patent on the TV scheme.] I had this idea. It was the practice at Bell Labs--we were always instructed to write things down and get them signed. So having thought about this a bit, I guess I wrote it down. But I never really worked on that.

Nebeker:

Do you think you talked about it with others there?

Townes:

I probably talked about it with others there, and probably nobody expressed great excitement about it, so maybe what I did was drop it. [Chuckling] I still do that in a sense. If I have an idea about something that I think is new, I think about it for a while and maybe do some calculations about it. If the ideas are complicated enough I might forget them so, I write them down. I usually don't get it signed unless I think it's a very important patent. But I just write it down for future reference in case I want to come back and think about it more seriously. Now every once in a while, I decide an idea is something that's close to my heart and something I want to do. Then I do it. Or maybe a year later I'll go into that field when I have time. So ideas come up like that. Frequently, in most cases, I don't have time to pursue them very far. But if they're close to my own field or something I can fit in, and I think it's very interesting, well, then, I might well do it.

Nebeker:

It seems that that would be one of the advantages of a place like Bell Labs. For one thing, there you can get somebody else's reaction to an idea like this. If other people thought it worth pursuing, there may be someone there who can pick up on it.

Idea sharing at Bell Labs

Townes:

I think that's quite right. There are a number of occasions when I felt I had ideas and I'd go down and talk with somebody that's in the field. Frequently they will decide it's not a good idea, but every once in a while, they feel that sounds pretty good. We will try it. In many cases then I don't pay a lot of further attention. They may go ahead and work with these things.

I remember one just because not so long ago I was reminded of this. John Whinnery who was trying sometime ago to start in the laser field, was looking at self-focusing in a liquid, shining intense lasers in a liquid or solid, but for self-focusing. He decided that some of the effect was absorption which was heating up the liquid, changing its density, and this was giving some effects. But absorption was small. I suggested to him, to put in a little dye to control absorption. Put in some dye and see just what it does. Well, he did some of that and recently he was thanking me for that. I said, that I never remembered it. He said, "You know, that was very useful to us." [Chuckling] It was a simple idea, but the conversation back and forth between colleagues is an important phenomenon of any kind of science or engineering. People talk back and forth. I get ideas from other people, and sometimes I give ideas to other people. I find that interaction very important, and I generally have ideas when I visit somebody else's lab. Some things I haven't seen for a while, and I may have a slightly different perspective on them. That generates some ideas. I think that's a fairly common phenomenon, but interaction is a very important phenomenon. At Bell Labs interaction was very good. Of course Bell Labs was a very big place. I interacted with all the people doing this kind of work, and some others that I happened to know. It was a good atmosphere from that point of view.

Nebeker:

I was talking a couple of years ago with Leon Lederman about a series of experiments he did at Fermi Lab. He made the comment that he's always been a great "telephone physicist." He'd call up people all the time to get help. That type of physics is a bit like a big engineering project. So it makes sense to have a lot of contacts, people that you can call.

Townes:

Sure.

Nebeker:

I wonder if you feel that your Bell Labs years helped you in building up these contacts.

Townes:

I would say that the first thing it clearly did was to give me some experience with electronics. The war did that for many physicists which produced a considerable burst of activity. But for me personally it was quite important in the fields that I subsequently worked in. So much of physics involves electronics now. In the 'thirties it didn't very much. Electronics was relatively simple, very specialized equipment then. There was very little complex electronics. Nowadays it's just all over the place. The second thing was the general idea of doing microwave spectroscopy and radio astronomy: that came partly out of that. But I'd been interested in radio astronomy before I went to Bell Labs.

Yes the contacts were important all those throughout my career. My contacts back at Cal Tech continued to be important. My contacts at Bell Labs were too. For some time I could call up people and ask them about things. I continued to have a good many contacts at Bell Labs, and many of my students have gone there. I would say it's been decreasing as the distance and time increased. Bell Labs did not react universally in a positive way to masers and lasers initially. That's because they weren't doing anything in the field. They recognized that the field of microwave spectroscopy was good physics, and though interesting, they thought, it would not lead to many applications. So while they supported me, they didn't want to build a regular group in the field. But I also didn't have the idea for the maser when I was at Bell Labs. That came a little bit later. When I began working on the maser at Columbia, my laboratory was open, and lots of people saw it, and they wondered what I was doing. They said, Oh, yes, well, that's interesting. Then they would pass on. Nobody was interested enough to really try to copy it at that time. As soon as we got it working there was of course more interest.

John Pierce of Bell Labs learned about it. He said, "Well, looks like you have negative resistance all right." [Chuckling] "Well, that's an interesting way of putting that succinctly, a professional engineer describes stimulated emission as negative resistance, which it is in a sense. "That's interesting." It was John who encouraged Bell Labs to hire Jim Gordon who had been the student who worked on the maser. So he decided that Bell Labs really ought to be thinking about this field. They hired Jim Gordon, and he went there. Then they later hired Ali Javan, who was a student of mine. And they hired Art Schawlow who was a postdoc with me. So I've had a close relationship there. Sid Millman, who was head of the physics department for some time (he was a good friend of mine), wanted me to be a consultant at Bell Labs, and he fixed up a very flexible arrangement. I was glad to do that because there were a lot of interesting things going on there. Arnold Penzias was another student of mine who went to Bell Labs. Now he is Director of Research. So I have a lot of good friends there, and I talk with them not infrequently about things.

Nebeker:

Well, for example, you asked John Pierce to be on the microwave Committee you formed.

Townes:

Yes, right.

Nebeker:

Is he someone that you could call if you had some question? I'm thinking of the early days.

Townes:

Oh, sure.

Nebeker:

What if you had questions about the properties of tubes?

Townes:

Tubes, sure. Noise also this. Sound wave tubes noise. Rudy Kompfner, too, I got to know later pretty well. Oh, yes, I would do that.

Nebeker:

Has that been very important to you, being able to call people or talk with people about things in their area of specialty, or not?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_19.mp3)


Generally that is important to me, talking with people, people who are knowledgeable, people I can call up. Bell Labs is a sizeable segment of that, but it's not the whole picture of all the various people that I do contact and call. But Bell Labs is a sizeable segment.

I might mention one other aspect that may be of interest to you. When I went on sabbatical in '55-'56, during the first summer I traveled around Europe a good deal. Then in the fall I settled down in Paris and worked at the Ecole Normale Superiere. There was a French physicist that I knew pretty well who did nice work. One of my students had gone over to work with him as a postdoc. When I got there, I asked him, "Well, what are you doing now?" He told me he was working on spin resonance, and he had found a material where the relaxation time of the spin was very, very long, perhaps a few seconds. You'd flip the spin one way in silicon, and then it would take a long time to go back in the other direction. The two directions have different energies, and this allows one to make a maser, a solid-state maser. So I said, Gee, that's certainly a very convenient substance for a solid-state maser, which would be tunable and would be a good amplifier, and why don't we try that out? So I tried it out there. However, he didn't have quite the right material or enough of it.

So during the winter I had to make a trip back to the States. I went back to Bell Labs, contacted my friends Jim Gordon and George Feher, a solid-state physicist who'd worked in spin resonances. He's since gone down to the University of California at San Diego. In any case, I contacted the two of them and said, I think I have an idea that Bell Labs would be interested in, that you may want to work on. But I need some material to work with in Paris, and I'll be there for about three months. I've got only three months to work in Paris, but if you could give me some material I'll work with it there and after I've left Paris, you're free to work on it and do whatever you like with it. They agreed. So I told them about the possibility of solid-state masers, and they produced this material for me, which I took back to Paris, and we tried it out. We never quite made it oscillate. We got some modest amplification, but we never quite made it oscillate before I had to leave. I went to Japan after that and followed my schedule. Then Jim Gordon and George Feher worked on it some at Bell Labs.

So there was a very close interaction in just this field, you see. I knew the people, and I trusted Bell Labs, and they were helpful. Now, it turned out that Bloembergen read about our work and had an improved idea, namely a three-level system. He suggested that. In the meantime, somebody at Bell Labs had also had this idea of a three-level system. They got wind that Bloembergen was interested in something that sounded awfully like this. So they contacted him, and they talked and agreed to give Bloembergen patent rights for use outside of the Bell Labs. They would get the patent, ownership would be Bloembergen's, but Bell Labs could use it free. The first one that was actually made was at Bell Labs, because they were already kind of primed in the business. Bloembergen was competing with them to make the first one, but actually Bell Labs made the first three-level maser system. Those are very concrete examples that collaboration and my own connection with the Laboratories was certainly important.

There's a little bit more maybe I might say about my interaction with Bell Labs. I guess my interaction with Bell Labs really has not faded very much because still have a good deal of contact with friends there. When I started working in the infrared region for astronomy, I was doing heterodyne work. Part of the things that I had been doing have grown out of the application of new technology to science. I've been close enough to technology to see some of the possibilities. Astronomers don't always see those things because they're frequently not awfully close to technology. But I am close enough that I can see the potentials. I tend to use new technologies and introduce them into physics and astronomy, and other fields, too. I was doing heterodyne detection of infrared, which was quite new for astronomy, using lasers and very fast detectors. Patel from Bell Laboratories has invented a kind of a spin laser which is tunable and produces infrared. He also invented the CO2 laser which we use as our primary source. But he also had a tunable laser. He came and worked with us a little bit, looking at the possibility of applying that. We tried it a little bit, but it didn't work too well, so we went back to CO2 lasers. I continue to get help by calling those people. They also, I hope, got a little help and a few suggestions from me from time to time. So there's a good deal of interaction.

Now the whole field of infrared astronomy is an interesting case because that really has been opened up by the efforts of the military and industry in developing technology, first with the development of solid-state technology and then with possibility of manufacturing and controlling things accurately. The military was interested in imaging the infrared and detecting the infrared. Industry, of course, has helped develop that. Now there is also a communications interest in the infrared, in the near infrared, at least. The detectors which were previously used were so primitive that there was relatively little astronomy that could be done with them. Now, they're wonderful. Very high quantum efficiency. I think the near infrared is perhaps the most fruitful area for astronomy, and a lot of astronomy is moving into that region. I, myself, have been working in the mid-infrared and the far infrared. Those use specialized detectors. The imaging detectors are getting to be very common in astronomy.

Right now I'm doing interferometry with two separate telescopes, using heterodyne detection in each telescope, and the mid-infrared is around 10 microns. For that we need very sensitive and very fast detectors. We want detectors with bandwidths many gigahertz wide. They're not easy. They've been developed in industry. Several industries have been very generous in giving me detectors to try out and use. There's a chap at Bell Labs now who's got a new kind of detector using quantum wells. It's clearly very fast and possibly quite sensitive, and he's giving us some samples and helping us out. That's not the only company that I contact. Santa Barbara Industries does work especially in infrared cameras and detectors. The General Electric people have given us detectors occasionally, too.

Nebeker:

Your mention of the heterodyne detectors makes me wonder if there was much transfer of technology from radio receivers to receivers of other parts of the spectrum.

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_20.mp3)


I think that this maybe is one of the relatively few cases. It's a very important case. There has, however, been a steady push to get down to shorter wavelengths with normal radio-type receivers, that is, heterodyne-detection receivers, pushing the microwaves down. Now industry has been interested in pushing it some, but science is interested in pushing it still more, down into the far infrared. Far infrared doesn't transmit very far in the atmosphere, so it's not of great industrial interest. But above the atmosphere it propagates and it's of interest to NASA, and it's of interest to scientists. And industry has been very helpful there in producing appropriate semiconducting materials or materials for superconducting junctions which are nonlinear and hence provide heterodyne detection. So there's a lot of interaction with industry. Much of the technology universities can't afford. We can't afford to have a big development laboratory for solid-state physics. Maybe if the government pays part of it. On the other hand, that's bread and butter for many industries. They develop these things and then very generously make them available for scientific purposes. In addition to that, whole new fields develop as a result of industrial breakthroughs. The solid-state work is a very good example. We use computers all the time, and we just couldn't do anything like the complicated work we're doing now without solid-state circuit elements. It's completely revolutionized the field, and so have computers. Infrared detectors are a somewhat more specialized thing, but that has been quite revolutionized, and I think infrared is going to be a very important part of astronomy in the next decade. Astronomers are waking up to that very rapidly now.

When you asked me about interaction at Bell Labs, yes, I still have really quite strong and frequent interactions with Bell Labs. But on the other hand, I interact with a lot of other people, too.

Nebeker:

Yes. I think that's a characteristic of a great many areas in the physical sciences these days, that research efforts are more often collaborative and not so often cases of one person in his laboratory working alone.

Townes:

Yes. We're dependent. Science and technology really have to develop together. That's a great disadvantage for countries that don't have a good technological base. Their scientists are at a considerable disadvantage.

Interest in radio, mechanics, and physics during childhood and college

Nebeker:

Could we change subject entirely and go back to your youth? I have a series of questions here that were raised by reading through some of these transcripts. You've said that you had an uncle who was chairman of the EE department at Clemson. What was his name?

Townes:

Dargan. D-A-R-G-A-N. Frank Dargan.

Nebeker:

Okay. You said that he gave you and your brother an old radio on occasion.

Townes:

Yes.

Nebeker:

Was there much contact there?

Townes:

Well, they were some of our favorite relatives, and we saw them with some frequency. He did not talk with us a great deal about engineering per se. We were as a family very good friends. He recognized our interest in gadgetry and said, "Well, I just bought myself a new crystal radio set, and would you like this?" He had a daughter, and I guess his daughter wasn't so interested. He'd give it to the two of us boys. And it was a crystal radio.

Nebeker:

Was it functioning at the time?

Townes:

Well, sort of. [Laughter]

Nebeker:

They can be temperamental.

Townes:

That's right. But we played with it, and he told us how to play with it, and work on it, and how to get the best receptions on it. So we played with it and enjoyed doing that.

Nebeker:

Was that just one of a great many things that you played with as a boy? Or were you especially interested in radios?

Townes:

I was interested in electricity and radios, and in high school I even tried to make a radio myself. A friend and I both wanted to make radios. I never really got it going properly. [Chuckling] He did. He got his going.

Nebeker:

Was this a crystal radio?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_21.mp3)


A crystal radio, yes. I never got mine going, but I wasn't a specialist in radio. I did a lot of different kinds of things. I liked to build things and make things. I lived on a small farm, and we did a lot of things ourselves. We took care of our own animals, and I milked the cows, and I did the gardening. I built houses for rabbits and pets and playhouses for other members of the family. We did all kinds of mechanical things, and I enjoyed that.

Nebeker:

You said that you took clocks and watches apart.

Townes:

That's right. My father owned a store which was rented to a man who sold clocks and watches. He knew this man, and this man gave him some of his junk, [Chuckling] clocks and watches, which my father was pleased about and brought home to us boys. We could play with them, take them apart.

Nebeker:

Did that become a major interest.

Townes:

It was just one of many, many things. We were very active in many things. My brother and I were both interested. We were very much interested in the out-of-doors and animals and plants and doing things of all kinds. Well, we boys in the family just liked to do things with our hands, to invent things and build things and make things. For example, I took shop, which was basically woodworking, when I was in high school. My brother had taken it, and well, we thought this was fun, and a good thing to know. It wasn't something that was required, but there was shop training there, so I took shop training. I took mechanical drawing. My brother had taken mechanical drawing, and, it was a good thing to know. So I took a course in mechanical drawing in high school. I was interested in technical things.

Nebeker:

I'm surprised that mechanical drawing was taught in high school.

Townes:

Yes. That was taught.

Nebeker:

You said also that as a youngster you read the Bell System Technical Journal because the local library received that.

Townes:

That was in college. The local library had that, and that was one of the few technical journals around.

Nebeker:

You said that you read a series of articles by Carl Darrow on physics.

Townes:

Yes.

Nebeker:

I've heard of those articles elsewhere, they were well received. Do you recall trying to read other articles in that journal?

Townes:

Yes, I remember looking at it in general. Every issue that would come in I would look at it. But I particularly remember Carl Darrow had written I think about three articles in three issues—two, at least—on nuclear physics, which was a new field then. Carl was an interesting case. Bell Laboratories had hired him from Chicago, a Ph.D., but he had some paralysis actually. He wasn't good with his hands at all. He had some kind of problem. It wasn't enormously noticeable, but it you watched, you could see that he couldn't manage his hands terribly well. Anyhow, he just couldn't do work in the lab, and the Bell Labs sent him to a meeting and said, "Well, okay, you go there and come back and report to us on the scientific meeting, what happened there." So he came back, and he made a wonderful report. So they started using him that way. [Chuckling] They said, "Okay, you study up on the latest science and come back and report." So he got in the habit then of trying to learn whatever new fields were there. He even wrote books on them, trying to summarize them.

Nebeker:

Do you know whose initiative that was to send him and then ask him to continue doing such?

Townes:

I don't know. By the time I came to Bell Labs, he was still there and active and busy that way. But he was well established.

Nebeker:

I don't know when Mervin Kelly became Director of Research.

Townes:

Actually I think he was a contemporary of Kelly's They were both from Chicago, I believe. So I would have thought Kelly might well do that, but on the other hand, he might have been too junior to do that at that time.

Nebeker:

Yes.

Townes:

Carl wrote up these various fields, and they were useful. To have a summary of a new field of physics written there in a journal was a great opportunity. So I studied those hard. He also wrote some things on gas discharges and also gas-discharge physics. I think I remember an article and also a book of his in this field.

Radio waves research

Nebeker:

You've also talked about the work you did on your spare time during World War II on radio waves from space. Looking at what we did yesterday, this device that you imagined for generating microwaves in inhomogeneous fields, made me wonder if maybe the fact that you'd been thinking about generating waves, may have played some part in your interest in figuring out how these waves were generated in space. It could be thought of as a kind of reverse engineering: How are these being produced?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_22.mp3)


Well, it's probably related, but maybe not quite so directly. I first got interested when I was in college, and I read about Jansky's work. I've never been able to locate that book. Some book mentioned that these radio waves had been discovered. Nobody knew how they were generated or where they came from. And I said, My, that's fascinating. I'd always kept that in mind, that that would be something interesting to figure out. So during the war, with what time I had available (and there were occasions when we weren't working and I could do some thinking about other things), I started thinking about that. I tried various schemes. Inhomogeneous fields may well have been one of the things that I tried and then threw out. Eventually I decided it had to be charged particle collisions that were doing it. I worked out the theory for that, and I showed it to my friends in Bell Labs. One of them was an x-ray man, and he said, "Oh, you know, it's something like that which Kramers worked out for x-rays." He referred me to the paper, and sure enough, Kramers had done exactly the same mathematics, worked it out for x-rays, and shown how x-rays were produced by charged particles colliding. It happened there was no difference except scale, [Chuckling] and so I had to give Kramers credit for it. But Kramers had never thought of applying it to radio waves, and this was the first time a really correct theory for radio emission from galactic ionized gas had been worked out.

So I wrote a memorandum. I know I wrote a paper and eventually published the paper. I also went down and talked with Karl Jansky and the people at Holmdel at some length. Karl was very interested, and I told him this theory about what I thought was going on. He had some data that he had never published. I was interested and asked him if I could use that and publish it and mention it. He said, "Oh, yes." He didn't feel it was definitive enough to publish as a separate paper, but that was fine if I published it. So I talked with Jansky at some length and really thought very seriously about going into the field. As it turned out, I didn't. I think you may have read the story about Professor Bowen whom I'd known at Cal Tech. He was very nice to me, and he's a very fine physicist. He'd become the head of Palomar and Mount Wilson at that time, and I talked with him about it. He said, "Oh, radio waves will never give you any information about astronomy." [Laughter]

Nebeker:

That's very interesting. One might write off such opinions on the grounds that astronomers don't know anything about radio. But his was a reasoned objection to it, that you couldn't get the resolution.

Townes:

Yes, that's right. As a matter of fact he had even helped a student do some work on radio waves. A physics student who did his thesis on radio waves on an ionospheric problem, I think. He was knowledgeable. It's completely logical that a radio wave is too coarse an instrument. You can't get any directivity, so you can't learn very much. That was the conclusion. Of course, he just hadn't gone far enough. [Chuckling]

Nebeker:

Yes. He was probably quite exceptional in understanding radio waves.

Townes:

Sure. He was a very reasonable person to consult about it. He was well accepted among astronomers as being very knowledgeable. I wasn't completely convinced about it. On the other hand, he didn't make any good suggestions as to what I should do with it. [Chuckling] I didn't know for sure what might best be done in the field.

Bell Labs approaches to radio astronomy and microwave spectroscopy

Nebeker:

I have a copy here of your internal memorandum arguing for working in microwave spectroscopy at Bell Labs. Do you think if you had really pushed for working in radio astronomy, that you might have had similar success at Bell Labs in being allowed to pursue it?

Townes:

Oh, yes, I'm sure they would have allowed me to work in radio astronomy. I think they would have felt it was somewhat related to the kinds of things they were doing.

Now there's an interesting kind of controversy: one of the things that Paul Forman said in his write-up of me was that I was turned down by Bell Labs and not allowed to work in radio astronomy. [Chuckling] I said, "Now wait a minute. I never even asked them." I feel sure that if I had asked them and written a memorandum as to why it was reasonable that I could have done it. Connected with that is some writing that Jansky was not allowed to continue in radio astronomy. Somebody in his family has apparently said that. I don't think that's correct. I talked with Jansky at some length. He never mentioned it to me! He encouraged me to do it, and he seemed interested, but said that he felt he had done as much as he reasonably could. I felt that he just decided that he'd gone some distance and that it was up to other people now. He was doing something else. He certainly never mentioned any dissatisfaction, any impediments to this.

I talked with Arno Penzias about it recently, too. He talked with a man named Crawford, I believe it was, who was there and who knew Jansky well. He said he never saw any indication of that. I think Bell Labs would have been quite open to that field. It would be fairly natural for them, and they were generally open. Radio astronomy, as you realize, is another field to which technology has contributed enormously. It was physicists previously engaged in radar during the war, who then really started the field. Physicists and engineers, a combination, really got the field going and invented the interferometer and various refinements. They were doing good work with it even before interferometry. They also started interferometry. In the U.S. there was nobody that took the lead to do very much. Australia and England were particularly involved, and then the Dutch.

Nebeker:

Do you have regrets that you didn't fasten on that field at that time?

Townes:

No. I don't really at all. microwave spectroscopy was a very good field, and it worked out very well. It was also integral to the laboratory. I could work intensively in the laboratory. I didn't have to go out and build a big piece of equipment. So I think it worked out very well. I think radio astronomy would have been good, too, but I can't at all say it would have been better.

Bell Labs working environment at Murray Hill

Nebeker:

I'm reminded of something I meant to ask earlier. Did you continue to work at the West Street offices of Bell Labs throughout your time there? I know you were away when you were testing.

Townes:

Yes. I was away sometimes. Then when Murray Hill was built I moved out there.

Nebeker:

Do you recall when that was?

Townes:

I don't recall exactly when it was. I remember it was being built before we were married, which would have been spring of '41 I visited Murray Hill, and some of the buildings were up.

Nebeker:

Was it before you started your microwave spectroscopy?

Townes:

Oh, yes. I moved out there during the war, about the time I started going to Florida.

Nebeker:

That was early '42.

Townes:

Yes, right. I meant the west coast of Florida. It wasn't before I went to Tampa. I might have moved there in the summer, or fall of '42. Somewhere around there.

Nebeker:

Did that seem a better environment to you?

Townes:

Well, it was a nicer building. It didn't have Greenwich Village nearby. [Chuckling] We ate in the cafeteria there.

Nebeker:

Did you and your wife move?

Townes:

Yes, we moved.

Nebeker:

You had a daughter by then?

Townes:

Well, yes. Our first daughter was born in June of '43. I was certainly going down to Florida by about then. My wife brought her down to Florida sometime a little bit later that year. We rented a house in New Jersey.

Nebeker:

In Murray Hill?

Townes:

No, it was in Mendham, which is not very far away. When I started going to Florida so much, we gave up the house. I came back and rented a place in Manhattan for two or three months and then went back down to Florida. I rented a place in Chatham for a while and went back to Florida. Then I rented another house in Chatham. [Chuckling] We went back, and then still another house in Chatham while we were there.

Nebeker:

Your wife took to moving?

Townes:

She was pretty flexible, and the children, too. I guess the last house we had in Chatham was shortly after the war. We had two children by then. So it was still moving around, but most of it was a wartime hardship. It didn't seem unproportional to what other people had to do. After all, the military was being shifted all around, and we were a young couple much like many military people. I didn't feel particularly oppressed about it. It was just something that had to be done.

Master's thesis on Van de Graaff generator; physics and engineering background

Nebeker:

To move a step back in time. Your master's thesis at Duke was with Woodbridge Constant with a Van de Graaff generator. You did work both with the apparatus and with analysis of its operation, is that right?

Townes:

Yes. I did some analytical work. I remember I found a factor of 2 error, at least a factor of 2 that had been mistaken or neglected in the previous theory. [Chuckling]

Nebeker:

How much work with apparatus was that?

Townes:

Quite a bit. Two Van de Graaff machines had been built, but they didn't work very well. What I had tried to do was to try to bring them up to good performance and make some modifications in belts and machinery, and the way charges were picked off the belts, and that sort of thing. Then I made tests on them. So it was clearly an apparatus kind of job. I never did any physics with them. I don't think any physics was ever done with them actually.

Nebeker:

I was interested in that because it might be regarded as part of your engineering background.

Townes:

Well, yes, it certainly was. I mentioned being brought up on a farm. I think a farm is a good place for both experimental physics and engineering. People have to make do with what's there. They invent things and make things and fix things. So I was pretty well experienced with that but not with quite such sophisticated apparatus. But I had no trouble with trying to make things work.

Nebeker:

At Cal Tech you worked with this gas-diffusion method of separating isotopes. Now was that an apparatus that Wooldridge had worked on earlier?

Townes:

Yes, that's right. It happened I had an interesting interaction with Wooldridge. He had left before I got there. But here was this empty lab, and I decided to work with Professor Smythe for various reasons. He was tough, and he didn't have any other students. I felt I would get a lot of attention. So he suggested I work on this in the empty lab. Wooldridge had left all the equipment there. He had separated isotopes, but had not been able to carry out any other physics with them yet because he had built the whole system. He'd overseen putting the system together and getting it working. I think some of the pumps were cracked, and there were other leaks in the vacuum system. A lot of things needed fixing up and improving. So I fixed them up and improved them, and enlarged it some. Eventually I separated isotopes and then studied the isotopes. It was a continuation of his work.

Nebeker:

But your thesis was more the study of the isotopes than the separation process?

Townes:

Yes. That's right. Doing Wooldridge's thesis, I would say, was building the apparatus and getting the separation. My thesis was perfecting that system, using it, and then doing the studies of the spectroscopy on the isotopes.

Nebeker:

You've talked about your taking Smythe's course and then helping him with his new book, solving all of his problems. I wonder if, in looking back on it, having done all of those problems gave you a facility with that kind of analysis that helped out at Bell Labs.

Townes:

It certainly did. His problems were always interesting. He liked to produce problems which if you thought through them well, you could solve relatively simply. But otherwise it would be hopeless. [Chuckling] He liked to make people think and try to see through problems. So I found him very instructive and interesting, and amusing. I like puzzle-solving, and it was a bit of a game as well as learning things. Of course I plowed through all of his book very thoroughly and corrected his problems. That was part of the reason for working them. He would give them to me to correct and be sure that they were right. So I knew that field. I think once you know one basic field very thoroughly, that's an enormous help in almost anything you do. You can apply that in many ways. It's wave theory, it's field theory. [Chuckling] You can apply it in all kinds of ways.

That's true of many other subjects. If you really learn them thoroughly in all of their aspects, you have a very powerful tool. And, yes, that has been very important to me. It was certainly one of the things that gave me a start at Bell Labs. It has to do, of course, with microwaves and microwave spectroscopy, and it had to do with radio astronomy. It touches on so many fields. I think learning that field as thoroughly as I did has really been quite important to my career.


Audio File
MP3 Audio
(143_-_townes_-_clip_23.mp3)


There's a more general aspect that I would comment on: I think it's very important to learn a subject well enough, to think about it enough, to think about it from all angles, so that you feel at home with it. You feel it's a friend. [Chuckling] Something you know intimately. If somebody asks you about it, you can express it back to them in your own way or anybody else's way. You can look at it from any angle and still understand it thoroughly. This gives one a distinctive idea of what happens, and what's going to happen if you're given a certain set of events. You just see it. You visualize what's going on. It's instinctive.

Now I always work out things with equations, too, to be sure it's all correct, to be sure I haven't missed anything. I usually think things through intuitively first, work it out with equations, and then think through the equations: Now, what do these equations really mean? Is that really what's happening? Do these equations really describe the right physical situation? I think the intimacy you have with a given field, is very important for quick thinking and also for exploring a problem thoroughly.

Teaching is another way of learning thoroughly, and I have profited from that. I think if you teach a field, then you have to look at it in any ways your students want to look at it. You have to look at all aspects of it. If you just learn it in other ways, you may well skip over certain parts you don't think are too interesting. If you've got to teach it, you have to look at all of those. And I have gotten some very good ideas just having to learn certain parts of a subject which I've otherwise not worried about, and I'll suddenly see something that's there that I'd missed. So I think teaching is another way of grasping a subject thoroughly and knowing it well enough so that you're at home with it.

Employment at Columbia University

Nebeker:

I have a question about your accepting a job at Columbia when Rabi offered you the job. I take it that you weren't actually on the job market at the time. You weren't applying to different places.

Townes:

No.

Nebeker:

I know you also preferred a university setting. But were you content to continue the work at Bell Labs?

Townes:

Oh, yes. I had expected to continue on at Bell Labs sometime. In fact, we had just bought a lot in Summit, New Jersey, and we were planning to build a house. We had the house all designed, and we were going to live there. [Chuckling]

Nebeker:

That's a pretty good indication.

Townes:

I wasn't looking around, but I really preferred a university atmosphere. And while Bell Labs was very generous, at the same time they had their restrictions, and I had to persuade them that what I wanted to do was a reasonable thing to do, and then they would allow me. At a university, particularly if you have a tenured position, you do whatever you want. Nobody says "no." They may not give you money, but you can go in any direction. [Chuckling] But I generally like a university because I like teaching, I like interaction with the students. I also like the breadth of the university, having people in the humanities and other fields that you can interact with. It's something I appreciate. But I had no immediate plans to leave Bell Labs at all.

Nebeker:

And it didn't look like they were dissatisfied with your working in microwave spectroscopy.

Townes:

No, they were generous to me in allowing me to continue. They said that they were pleased to have me continue the field, but they just didn't think it had enough application to communications to make it worthwhile to hire another person. I had propositioned them about expanding the field and hiring another person or two in the field, and they felt they didn't want to do that.

Microwave spectroscopy

Nebeker:

You've mentioned that at this time Good at Westinghouse, someone at RCA, and someone at GE were working in microwave spectroscopy but those companies didn't feel that that was worth their investment.

Townes:

That's right.

Nebeker:

That I can understand. What I wonder about is why the companies supported that work in the first place.

Townes:

After the war, it was the commercial electronic companies that had equipment of this type. People there had been working with microwaves. Those companies and Columbia University and MIT had microwave equipment also. So they had had specific laboratories connected with the radar effort. So it was people in the radar effort basically that started into the microwave spectroscopy with equipment they already had on hand. That's how it started at RCA, Westinghouse and General Electric.

Nebeker:

That's a case of individuals making use of the technology.

Townes:

That's right. They saw something that looked interesting, and they thought they would try it, and they did it, and they published a paper and got a little notoriety from it. The company, of course, was pleased that people were doing something good. But on the other hand, after a few years, they decided, well, okay, that's nice. But that's not really the thing that we ought to be supporting.

Nebeker:

But it suggests that those scientists and engineers at those places had a little more freedom than people might imagine at those companies.

Townes:

Yes.

Nebeker:

They could pursue what they saw as something interesting right after the war.

Townes:

Yes, they certainly could for a while. And I think even after they stopped, that general atmosphere continued. That, yes, people could try new things and try them out. But in the long run, how much of that they supported was the question.

Nebeker:

Sure.

Townes:

At RCA the chap who was there left and went to UCLA engineering school to teach there and to continue to try to do some of this kind of work. My understanding is RCA discouraged him from continuing in this field. And then he got an offer from UCLA.

Now General Electric I know very well. The chap there told me that he was just told he'd have to stop. That he should do something that was useful to the company using his microwave knowledge. So he used microwaves to measure the properties of solids which they needed in their dielectric materials. He was told that he should do that. He was really rather broken up about it. He had worked in the field for a few years and was good at it.

I think in the case of Westinghouse, it faded than was shut down. The people did what they could do and were beginning to lose out. They were initially some real leaders in the field, and then they began to lose out and they weren't so much leaders, and they lost interest themselves, and went on to other things. I think Westinghouse may have encouraged them to change but didn't make a dictum about it.

In the case of Bell Labs, they said they would be pleased to have me continue it. It was nice work, and they were glad to see it. But they just didn't feel they ought to enlarge it.

Nebeker:

Right. When you left, no one picked it up?

Townes:

No one picked it up. They didn't feel it was important enough to them to hire somebody to do it.

Millimeter Wave Committee of physicists and engineers

Nebeker:

I have a question about the Millimeter Wave Committee that you organized. I don't have all the names of the people on it--Whinnery and Pierce and John Strong and Marvin Chodorow. You say in the long interview that these people were basically electrical engineers.

Townes:

They were. Chodorow had started out in physics but then switched to engineering. He's been an EE professor at Stanford. Whinnery and Pierce were engineers from the beginning. Strong was a physicist. Strong was in infrared. He's a physicist, an experimental physicist, in the infrared. There was another physicist, who was in superconductivity at Ohio State. I can't think of his name at the moment. He was British. And there may have been one or two other people. My idea was to base this on clever engineers who had worked with microwaves, but to put in people in different adjacent fields because these adjacent fields might contribute and there might be some ideas from these areas that would come up. For example, somebody in the infrared might find a way of producing a continuous oscillator. Superconductivity, was a new field. I felt it had potential. Maybe that it might generate some energy. I think I had somebody in solid state also on the committee.

Nebeker:

So the idea was to have both physicists who might bring some entirely new idea to it, and the engineers.

Townes:

The very experienced engineers. That's right. Those who really were right in the center of it and would be good judges of what was being done. I think there was a chap from Hughes I had on it, also. An engineer who had done some very original work in short waves. I can't think of his name right now.

Nebeker:

Well, that answers my question about the composition.

Townes:

What the mix was.

The role of logic in research

Nebeker:

Yes. I was fascinated by quite a few examples in your career of cases where someone has used general principles--or even more specific analysis--to prove that something wasn't possible. Some of the examples are that Pierce and Kompfner showed that an amplifier, that is one using electron tubes couldn't be much better than the present amplifiers. And your showing that the Second Law of Thermodynamics showed that you couldn't get much power out of these. And Bush arguing that rockets could never take a man to the moon. And I think there were a couple of other examples. [Laughter] What comments do you have there?

Townes:

Well, I think that's an important phenomenon, and I'm rather struck by it, too.


Audio File
MP3 Audio
(143_-_townes_-_clip_24.mp3)


We use logic the best we can to try to block out what directions we should look in and what's possible and what's not possible. If we can lay down general laws that sort of clear up one particular area, that say you can or can't do this, that's very helpful. So that's a very important part of our reasoning. But frequently we get trapped in being incomplete in that logic. There are also emotional biases that tend to blind you so you lay down general principles that kind of fit your emotional biases without thinking about possible exceptions or possible boundary conditions. You're not really necessarily looking at the right boundary conditions. You may be very logical about a particular set of boundary conditions which you think are the reasonable ones, the ones everybody accepts. So, yes, you're completely right. But they aren't the right boundary conditions, necessarily. You can find ways around them.

That frequently happens in the breakthroughs in science—or engineering, for that matter. Now of course if you have some confidence in the nature of the problem, as they thought of course when you amplify radio waves with electrons, with electron beams. The interaction of electrons is the way you get energy from the electrons and the radio waves. And so they treated it as generally as it could be treated probably with that kind of mechanism. But they didn't think about other mechanisms. They probably were quite unacquainted with stimulated emission as a matter of principle. Most engineers were. As a matter of fact, many physicists, while they'd been taught it, didn't think about it very much.

Now I don't know whether you noticed in any of these tapes that after we'd gotten the maser going and recognized that you can get amplification with this basic quantum noise. There was a chap down at JPL who came to me. He said, "I'm wondering. Can't you do the same thing with a parametric oscillator?" And I said, "I think you're probably right." He said, "I've been working on parametric oscillators and thinking about this, and it seems to me you could do the same thing and get essentially a quantum noise kind of amplifier." And I said, "Well, yes, I think that's right." It woke him up that here is another kind of device that people had just been neglecting, the nonlinear interaction and parametric amplification. Of course those are used more as amplifiers now than masers are. They're just handier.

He was stimulated by the fact that we had done something that showed that you could do this. Then he'd been working on parametric amplifiers, and he said, "I believe those maser arguments apply to what I'm doing, too." And they did. So the parametric amplifiers came along. Those were another class of amplifiers which people had just not been thinking about, you see.


Audio File
MP3 Audio
(143_-_townes_-_clip_25.mp3)


Another area which I've also been involved in is molecules in interstellar space. When I started, I proposed that you could find stable molecules in interstellar space. At least, I thought we ought to look. I thought they might be there. I was asked to give a talk about this to an international astronomers' meeting. People congratulated me, thought it was very interesting and that we really ought to do this. But nobody did it. Finally when I came here about five or six years later to Berkeley, I decided, "Well, I ought to try to do that. I think people are neglecting that." The astronomers all argued strongly against it, saying, "The density in interstellar place is just not high enough. And the ultraviolet's going to destroy the molecules, and they can't be there."

Now there were some papers which I found about dust clouds, stating that nobody'd found any hydrogen in the dust clouds. There's just dust apparently. No gas. Why was that? And one paper said, it possibly might be hydrogen molecules instead of atoms, and that's why we're not seeing them. But nobody paid any attention to that. Nevertheless, that bolstered my thinking that that's very likely the case. People just weren't seeing these things. So we looked, and there they were. Since then it's suddenly become a very important field of astronomy. The former head of the department here mentions this in a book he wrote, he mentions that he told me it wouldn't work. [Chuckling] He had two very good theoretical reasons for it.

Nebeker:

Right.

Townes:

Now his reasoning was completely logical in terms of what was assumed about interstellar space. What was assumed, this low density, just wasn't so. I think that's rather typical of discoveries and change. We block out things we think we know, and it's important to do that so we know what directions to look in. But then we close our minds a little too much as to how generally those arguments apply.

Nebeker:

Sometimes it's an outsider who doesn't really understand the reasoning against it.

Townes:

That's right. Or who's looking at it a different way. Or who's just stubborn or something, that sees this. There's another case in this molecular thing. When we found ammonia, Norman Ramsey from Harvard said, "Ed Purcell kept me from finding ammonia before you did." [Chuckling] I said, "How was that?" Ed Purcell, by the way, is the first person who found radio radiation from hydrogen atoms in interstellar space, so obviously he's an expert and a very good physicist. He's an excellent person, who knows a lot and is very smart. Ramsey said, "Well, I was going to look for ammonia, and I had a student who was going to do it. And then Ed talked him out of it, saying it couldn't possibly be there. You couldn't possibly see anything." [Chuckling] And he had good reasons for it. He explained, very quantitatively, just why it couldn't be there. [Chuckling] Ramsey said, "I wasn't convinced, but the student was, so I couldn't do anything." [Laughter]

Nebeker:

Well, it's probably wise to be cautious as a grad student.

Townes:

You have to take some chances and be a little crazy.

Edwin Howard Armstrong and quantum electronics

Nebeker:

I have another topic entirely. You mentioned that you had some acquaintance with Edwin Howard Armstrong at Columbia.

Townes:

Oh, yes.

Nebeker:

He's getting more and more attention these days. Many engineers regard him as one of the greatest engineers of this century, and I'm curious what you recall of your interactions with him.

Townes:

I always thought that he was quite remarkable, the number of different things which he did. He was very imaginative.

Nebeker:

I was wondering more about your personal contact.

Townes:

I used to see him at lunch fairly frequently. He was a nice person, interesting. I used to talk with him. He, at that time, was right in the middle of a patent suit and used to talk a lot about that. He was very annoyed at RCA and other people who wouldn't recognize his patents. Particularly frequency modulation. RCA was claiming that frequency modulation occurs in nature, which then rules out the patentability. Armstrong was arguing it never occurs in nature. That didn't seem quite right to me, but it is not so common in nature. He felt very strongly about this, but he was not too unreasonable. I enjoyed having lunch with him.

Nebeker:

He could talk about other things than his patent suits?

Townes:

Yes, he could.

Nebeker:

I know that became his main occupation in later years.

Townes:

That's right. He had a lot of patents to fight. By then he'd spent a whale of a lot of money fighting, and he still wasn't winning. But we talked also about general things. He frequently ate lunch with the physicists at the Faculty Club. We'd meet over there. So I would see him with moderate regularity.

Nebeker:

He was interested in what was going on in physics generally?

Townes:

I guess he was probably more interested in my work than in others because it had something to do with electronics. He didn't particularly talk about high-energy physics or nuclear physics, but he talked about electronics. I was very interested in him because he was such a character and had done so much.

He may have influenced me some in the following respect. He eventually jumped off a building and killed himself, and I think some of it was family problems. But I'm sure he was very depressed about this RCA situation, and he'd spent much of his fortune on it.

When the maser came along, I was supposed to submit the patent case to people in the university so they could patent it. Which I did. They didn't have any patent policy, so they called together a committee to decide what to do. Major Armstrong had always been just patenting his own things without going through Columbia at all, and he had set a precedent. According to the agreement that I had signed for the support of my lab work, the patent belonged to Columbia University and I was obliged to turn it over to them. They finally decided that they really didn't have any policy, and that if I'd like to patent it to go ahead and patent it myself. Columbia wouldn't bother about it.

So I did that. But knowing Armstrong's problems and how difficult these things could be probably was part of my reasoning. I don't remember it that clearly. In any case, I decided that I didn't want to get locked up in a lot of patent problems and spend my life that way. The Research Corporation was handling patents and giving the resultant money to universities, and they'd given substantial amounts to the physics department. So I gave the maser patent to the Research Corporation and let them take it over and worry about it and give me a certain fraction back. Then I didn't have responsibility for it, and I could forget about it. So that's what I did. [Chuckling]

Nebeker:

That was very wise. I take it that you coined the term "quantum electronics" for this conference in '59.

Townes:

Right. The Navy suggested that we might have a conference. They would pay for it, but they asked me to organize it. So we got a committee of people together. The name was picked out over lunch. I said, "The field is getting big enough. It really ought to have a name for it. So I suggested "quantum electronics." Everybody liked that so we called it quantum electronics. [Chuckling]

Physics in electrical engineering curricula

Nebeker:

So that's where that started. You probably know that ever since electrical engineering was an academic field, there have been debates about how much physics should be in the curriculum. It's a pendulum that swings back and forth with demands for more practical and more applied work, and then calls for more grounding in physics. In the long interview you commented that you thought that MIT had gone a little bit too far toward industrial work in engineering with some neglect of science. You saw it as part of your job there to build up science there. I'm sure that you were talking about the school as a whole there.

Townes:

Yes.

Nebeker:

Did you feel that about the engineering curriculum in particular at MIT?

Townes:

No, no. I wasn't arguing that the engineering curriculum per se was wrong. Rather I argued that the general emphasis of the school and its abundant contacts and commitments to industry had distracted them too much.

Nebeker:

Yes. I understand that sort of conflict of interest for faculty members.

Townes:

That's right. While MIT had a very, very important engineering school, in order to do the job well even in engineering, they ought to be cultivating very good science.

Nebeker:

So your effort there was to build up the science separate from the engineering.

Townes:

Just in addition to the engineering.

Nebeker:

Right.

Townes:

I also wanted to cut down the conflict-of-interest problem, which was pretty severe. Department heads, for example, would be also running companies on the side, or serve as the president of a company, or something like that. We stopped that. The engineers didn't all like it, of course, [Chuckling] because it limited their freedom. But I felt that the curriculum in engineering was appropriate and fine.

Nebeker:

So you didn't see any need to get more physics into the EE program, for example?

Townes:

No, no. I think they did pretty well in that respect, actually. They were fairly close to the physicists.

Impact of defense spending on Cold War

Nebeker:

This is a very general question. As you look back on your career, I'd like to know about your attitude toward the defense work that you've done, especially on these advisory committees. There are people who argue that the strength of the United States in the postwar era was decisive in bringing about the demise of the Soviet empire. On the other hand, there are people who argue that this recent demise shows all along that we were wasting a lot of resources in defense spending, that we were much more worried than necessary and putting too much into that. How do you regard this?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_26.mp3)


Well, I think we have wasted some money here and there. I think the SDI, for example, has been a bit wasteful. Nevertheless, I do think that's been an important component of the Soviet demise. Really, their only strength was military. That's obvious now, but that seemed pretty plain to me quite early in the game, that otherwise it was a second- or third-rate country. It was in the military where they were outstanding, and it was because they were putting so much effort into it. It's clear that their ambitions were such that they would continue because this made them an important nation in the world. They would continue until something else happened. I felt we had to be well prepared and see that we provided some kind of checkmate to that.

There was an argument for some time that stated, let's just force them to spend too much and then they'll collapse. Most people look down on that. That's just a perverted kind of an argument. You mustn't ever think of trying to do that. That would be silly. Because it will cost us money, cost them money, and it'll never work. They'll never give up. As it happened, that's about what took place. That's one of the things that took place. There were various other reasons they failed, but certainly their excessive spending on the military was part of it. They could have relaxed on that if they hadn't felt they had to keep up with us. The SDI, I know, had them very worried. It was not because their scientists felt the SDI would work. I was talking with their scientists quite extensively, and they were convinced, as I was, that the SDI would not work as Reagan said it would. Rather, they were afraid that this would put us so far ahead technically in the military field that they couldn't compete. They were very concerned about that. They were highly impressed with U.S. technology and very concerned. They were much more impressed with it than we had been. [Chuckling] They felt they were behind, and they were very worried. I think that pressure certainly contributed to their collapse. Whether they would have collapsed otherwise is debatable.

I think it's clear they were dangerous. They had weapons which are really very extensive, very powerful, and very successful. They were very good in space and military work. They put all their best people in those fields. So they had a very powerful attack force. If they had wanted to use it, it could have been disastrous. I think we had plenty to worry about. Now we might still have cut down our expenses some. Overall I think it was important for us to see that they realized they could not overcome Europe or the rest of the world without enormous losses themselves. I think that was very important.

Physicists'  and engineers' influence as government advisors

Nebeker:

You know that World War II has been called the "physicists' war." Thousands of physicists served in an engineering capacity during the war. Nevertheless, it may appear to the outsiders strange that physicists have retained great influence on government and military advising. Others think that it's the engineers who are able to advise about feasibility of projects. What's your view? Do you think that physicists have had perhaps more influence than they should have in government advising? Or is it appropriate?

Townes:


Audio File
MP3 Audio
(143_-_townes_-_clip_27.mp3)


Well, I think that during the first period of PSAC (the President's Science Advisory Committee), they were effective. But as time went on, they became less effective because certain political elements began to come in, and the question came up of whether or not they were dispassionate in advising the president. Of course that's the PSAC group, which was a very important group. A number of people asked me why I or people like myself should be asked about policy matters. I point out that technical people get asked about matters involving the government because the politicians realize there are some things they don't know. If you talk about problems in social science, the politicians think they know that and that they don't need to consult anybody. They'll consult their buddies, other politicians. They think they understand those things. But when it comes to highly technical matters, they know they don't understand it, and they've got to call in somebody. Now when they call in somebody, obviously policy matters and details are involved. So this is why technical people get involved more than it might seem that they should in terms of their general knowledgeability about society. Now so far as the engineering is concerned, why is it that physicists have played a larger role in advising.

Nebeker:

I'm thinking of people like Feynman, who was in the group investigating the Columbia disaster. It seems it should be engineers rather than physicists. Somehow the physicists get asked very often.

Townes:

Well, Feynman is a special case. Feynman was quite a showman. He served on virtually nothing else. He was unwilling. When he got asked to do this, he did it, and he did it with some flair. He's a good showman, and he's a smart guy. I think in part the physicists get asked, and maybe should, because if you look at the future technical developments rather than the past ones, the physicists may have a better grasp of what could be coming up because they're working in a more fundamental field. If you're going to look at the future, then that's not so unreasonable.Actually a lot of engineers do get involved in these committees and studies. Certainly a lot of physicists do, too, and maybe more than a reasonable balance. I think another reason, though, is that a lot of the engineers are in industry. Generally, I think it's not impossible to say that the best engineers are in industry, the best physicists in the universities. That gives the physicists a certain kind of independence in principle at least. Industry has an axe to grind. They have conflict of interest. If you look at the committees, you get an industrialist, he's probably head of his company if he's a good engineer. He's probably got commercial interests in what the government's trying to decide, and the government shouldn't call on him. They should call on somebody outside, maybe a different kind of a company. But if it's a different kind of a company, then perhaps the person isn't quite so close to it. I think that's another thing that tends to rule out engineers.

Now the recent administrations, though, actually tended more toward engineers. Partly for the reason that they're more business-oriented, and partly for the reason that the academic people tend to be more Democratic rather than Republican. Getting engineers to advise may also be a question of reasonableness and balance. The head of the new Space Advisory Committee, for example, is an engineer. A lot of that committee is pretty heavy with engineers. I think there's a tendency to go in that direction now.


Audio File
MP3 Audio
(143_-_townes_-_clip_28.mp3)


World War II, of course, did wake up the country to the world of physics because physicists did contribute heavily. Many of them were largely doing engineering. But they did bring in some fresh ideas, and they learned some new things themselves. So physics became very popular. Perhaps that public image of physics is part of it, too. I think it's the combination of the public image and the fact that most of the best physicists are in academic life and presumably don't have the kind of conflict of interest. Plus because their work is more basic and forward looking, their background is a little more applicable. Maybe those are the reasons. Many times it's just a question of judgment. Experience and judgment. If you talk about how do we build this piece of equipment, then clearly the engineers ought to have better judgment. I think generally for that kind of thing the country does call on engineers. If you try to get a committee together to look at the future, then you're more likely to get basic scientists. That's my view. I don't pretend to be very profound on the subject. [Chuckling]

Institute of Radio Engineers and the IEEE

Nebeker:

It sounds very reasonable. I wanted to ask you about your connection with the Institute of Radio Engineers and then IEEE. Do you recall when you joined?

Townes:

I don't. [Chuckling] It's been so long that I don't really remember. I was reading their journals, certainly, from time to time, and I'd publish there occasionally.

Nebeker:

Right. I know your 1959 article on the maser amplifier for radio astronomy appeared in the Proceedings of the IRE.

Townes:

Yes. I published there occasionally. I'd read their journal some. I might well have just decided that I ought to subscribe and be a member. I don't presently remember, sorry to say, just how that came up.

Nebeker:

But you have been a member for quite a while, it appears since the 'fifties.

Townes:

Yes, I've been a member for quite a while. That's something I automatically would do. Once I joined I would just continue the membership.

Nebeker:

You said that some of the publications were useful to you.

Townes:

Yes.

Nebeker:

The IRE, or one of their groups, was involved in that first conference, the '59 conference on quantum electronics.

Townes:

We certainly had some engineers involved in the meeting. I don't remember an engineering group particularly. I don't distinguish very sharply between the kind of engineers that are fairly fundamental and academic and the physicists who are somewhat applied. I don't think there is a very sharp distinction. So at a meeting like that I would invite some physicists, some engineers.

Nebeker:

Have you gone to conferences run by the IEEE?

Townes:

I have in the past. I haven't been to one in a long time now. But then I haven't been to an American physics Society conference in a long time either. I largely go to astronomical conferences now.

Nebeker:

Let me ask about that '59 article on maser amplifier for radio astronomy. Why did you decide to publish in the Proceedings of the IRE rather than in some astronomy journal?

Townes:

The astronomers weren't interested and probably wouldn't have taken it. [Laughter]

Nebeker:

There were just so few then that would have understood it and appreciated it?

Townes:

Yes. That's right. No, I doubt that they would have published it. So IRE seemed to be a natural place.

Nebeker:

I took a look at a book, The Evolution of Radio Astronomy by J.S. Hey. This book lists important papers, and many of those were in IRE journals. One could infer from that--and I'm asking you if this is correct--that in a sense radio astronomy came out of the engineering community.

Townes:

Yes, it certainly did. Jansky was an engineer, and the engineers were pushing it. Grote Reber was next, and he was an engineer who built his own system in his backyard. Southworth, another engineer at Bell Labs, did some work. Yes, radio astronomy certainly came out of engineering. There were physicists involved, but they got involved in radar and were closely associated with electrical engineering. It was sometime before astronomers appreciated the field.

Nebeker:

In an article you wrote for Science called "Quantum Electronics and Surprise in the Development of Technology" in 1968, you talk about the connections between physics and engineering, and you write, "It was the drive for new information and understanding in the atmosphere of basic research which seems clearly to have been needed for the real payoff." I understand that point, and it's a very important one. You need people who are interested in fundamental understanding, and they're not always oriented toward devices or results. The other side of that is that it's very important that physicists and the ones doing that kind of fundamental research, are alert to the possibility of application. A clear example is from your Bell Labs notebook where you propose a new type of television tube. You got the mental gears for invention going. You saw that here's something that could be turned into a device.

Influences of World War II and Vietnam War on scientists' collaboration with government and industry

Townes:

Yes. Oh, I think it is very important.


Audio File
MP3 Audio
(143_-_townes_-_clip_29.mp3)


In fact I think one of the strengths of American science and technology that came about after World War II is that people had been mixed together, gotten to know each other and each other's fields. When they went back to the universities and into industry, they had that kind of background. I think that interaction was a very important one, and the interaction with government.

It was cut off by the Vietnam War. The Vietnam War made university people try to shy away from business and industry and shy away from government. I was here at Berkeley during much of it, and, oh, people would really jump on me for having anything to do with industry or the military. For example, when I went on the Board of General Motors, I knew there was going to be a lot of criticism. I contacted the president of the university and said, that the Chairman of the Board of General Motors asked me to form this advisory committee. It's a big company. At that time it was about 3 percent of the total GNP of the U.S. They felt they had difficulties and needed some advice. I felt it was a sensible thing to do. But I recognize that it might be criticized by the university, and I asked what he thought. Did he think I ought to do it? He said, "Well, I think on balance you ought to do it." So I did it. But sure enough I got jumped on in the student newspaper, the idea that I would have anything to do with a big commercial company like that. That was sinful. Nowadays it's welcomed. [Chuckling] People want it because they think maybe GM can give them some money. But then it was very bad, and so was my connection with government. I was compared with Dr. Strangelove in various ways. I would sometimes advise the military. I felt it was important for people to be in contact and try to see that things went in the right direction. There was a very bad dichotomy at that point. Now I think we're pulling back together again, but I think it's very important to find ways of producing interaction and the exchange of ideas and people who are familiar with both fields.

I think I've indicated in those articles that I felt it was my own fortunate combination of engineering concepts and physics together which was able to bring this about. On the other hand, how the fundamental work comes in is that industry was not all that interested in getting into very short wavelengths. They still don't do very much with millimeter wavelengths or sub-millimeter wavelengths for industrial purposes. To me it was quite important because there were scientific goals there. So it led me to push in directions that industry wasn't pushing in at that time. That's the sense in which fundamental work opens up your eyes and pushes you in different directions to explore new things. But I generally feel the interaction back and forth is very beneficial. One can maybe overdo it by being just a jack-of-all-trades. But on the other hand, I think that there must be some contact interaction by both to see that there's not a big gap between the academic work and industrial work.

The maser and coherent radiation

Nebeker:

That reminds me of a question that I meant to ask earlier. You've written, in considering the origins of the maser, that people who had done some work related to that earlier hadn't appreciated the fact that radiation would be coherent.

Townes:

No indication.

Nebeker:

There was no indication of that. I'm wondering why that's an important point. It's obvious that with some of the applications, such as the detectors or generators, it is important that the radiation be coherent. But one could imagine applications such as intense light source or something like that where it's not important that the radiation be coherent. Is your point simply that maybe the most obvious applications depended on the coherent feature and that that wasn't sufficiently appreciated?

Townes:

Amplification, for example, depends on coherence. Oscillation depends on coherence. If you want just a lot of light, that's true; you don't have to have coherence. But then that's not nearly as flexible or as interesting. I think most people who suggested this or realized it could be done, thought of it just as a kind of amusing trick. That yes, you could generate light this way. They didn't think of it in terms of high power or usefulness.

There was one exception. Von Neumann wrote a letter to Edward Teller suggesting that he might bombard semiconductors with neutrons, and that they would fall back, and there would be stimulated emission, and you would get a lot of energy coming out. He wanted a kind of blast of energy coming out of semiconductors. Interestingly, Teller never answered his letter. I asked Teller. He wouldn't comment. [Chuckling] He probably didn't remember a thing about it. Von Neumann really had that idea. It's clear he didn't think of the coherence or directivity. But he was the only one that I know of who really thought of getting some energy rather than just sort of saying that it could happen this way. Which is what was the case with most of the others. They realized the principles. I think I have said, I believe the smart scientist, if asked, "Would this be coherent?", would think about it for a while and then say, "I guess it would be." But it never crossed their minds particularly.

Nebeker:

I think that's an important point.

Townes:

Feedback was never suggested by any of the others, so that in terms of a relatively small effect, given by one pass only and no reflection, you might expect some emission and some effects. This is part of the reason that they never really took it all that seriously. There were some people who tried it, and then didn't see the effect and gave up.

Atoms and molecules as circuit elements

Nebeker:

I wanted also to ask about your statement in this internal memorandum at Bell Labs, where you say that atoms and molecules might be used as circuit elements as one got to shorter and shorter wavelengths. I can understand that with the maser as an amplifier or wave generator. Were there other circuit elements that you were thinking of?

Townes:

Yes. I was specifically thinking of other circuit elements because, as indicated, [Chuckling] I did not think of this as a way of generating energy.

Nebeker:

You were thinking of passive circuit elements?

Townes:

Passive circuit elements. The reason is the following. I don't know whether you know the similarity between the Bode Theorem and the Kramers-Kronig Theorem. This is something that had struck me. The Bode Theorem is about feedback amplifiers and the relations between phase and attenuation, the fundamental relationship between phase shift and attenuation in circuits. If you have a certain phase shift, if you measure all the phase shifts, you can predict the attenuation and vice versa. Bode worked out those relations. It turned out Kramers and Kroning had worked out those relations a number of years earlier for optics for the change in the phase or the wave going through something and its attenuation. They'd worked that out some years before. Bode didn't know it. People at Bell Labs didn't know it. I ran into that paper, and they just connected.

Nebeker:

The same mathematics?

Townes:

Same mathematics. Same principles exactly. If one is right, the other has to be right. People never realized that. They didn't think about it. Now I became very aware of it, and, in fact, I pointed it out to the people at Bell Labs who were making some mistakes because of that. They didn't recognize that parallelism. Now this makes a parallelism between circuitry and phase shifts and optical properties. And they also tell you that if you have a resonant response, that's going to shift phase in a certain way. It's going to have a certain attenuation, too. An atom or molecule absorbs, so if you send light through it or a radio wave through it, it's going to shift the phase; it's going to attenuate in certain ways. You can vary that attenuation and phase shift by varying the characteristics of molecules, varying the pressure, varying the fields that are applied to them. So you can even modulate those passive elements. That was the general idea. As you have to make things smaller and smaller, it may become difficult to make circuit elements by hand. Then you ought to start using resonances of atoms and molecules perhaps. By the time you get to optics, that's what you in fact do.

Nebeker:

Are there in fact today passive circuit elements in that wavelength?

Townes:

Let me give you one example to illustrate that. Suppose you try to stabilize an oscillator. You stabilize an oscillator on a resonance of something: a quartz crystal maybe, maybe a microwave cavity. A microwave cavity has a certain resonance. So it's a circuit element, and you stabilize the oscillator on the resonance of that circuit element. Now you can also stabilize an oscillator on a spectral line. That's the same thing. The spectral line does exactly the same thing. In a sense that's a passive element you're just using to mark a frequency in a resonance. So there is that.There is also material which you use to, let's say, focus radiation. You can build a lens, after all. You can build a lens for microwaves, make a plastic lens, let's say, for microwaves. That's been used some. That lens focuses or deflects the microwaves as a result of its resonances, even though they're not highly resonant. Still there are responses there which allow focusing. It's a phase shift just like you might have a batch of delay lines on an antenna, let's say, and do things that way. I'm sure there are some cases where attenuation is used. Certainly if you talk about circuit elements for laser light, then you're going to use attenuation and absorption and phase shift due to them. So there are some of those. I think if you ask about classical frequencies for electrical engineering, such as 1 centimeter and longer wavelengths, then the only things I can think about are this use of dielectric focusing and frequency stabilization. You get into the shorter wavelengths, you get into light, why then you're essentially always using atoms and molecules, for modifications of the light.

J. R. Wilson

Nebeker:

A final thing. You were wondering yesterday who J.R. Wilson was?

Townes:

J.R. Wilson.

Nebeker:

You worked first with Llewellyn. He was the second person you worked for that you couldn't remember.

Townes:

Yes. He was head of the tube lab. Sears, I remember, was another person I saw something of who worked the tubes. Oh, yes, Mendenhall was there. Yeah, I remember all those people.

Nebeker:

This was in May of '39.

Townes:

I see. [Chuckling]

Nebeker:

I'd like to thank you for the interview.