History Home | Book Catalog | International Catalog of Sources | Visual Archives | Contact Us

Oral History Transcript — Dr. David Wilkinson

This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.

This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.

Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.

Access form   |   Project support   |   How to cite   |   Print this page


See the catalog record for this interview and search for other interviews in our collection


Interview with Dr. David Wilkinson
By Martin Harwit
In a small conference room in Jadwin Hall, Princeton University, Princeton, New Jersey
September 27, 1984

open tab View abstract

David Wilkinson; September 27, 1984

ABSTRACT: Family background and early education; undergraduate studies in engineering at University of Michigan, faculty who influenced him; doctoral thesis at University of Michigan with Richard Crane on the g-2 experiment. Postdoctoral instructor; decision to work with Robert Dicke on gravitation at Princeton; funding in the Princeton Department, faculty, the ongoing NSF grant for Gravity, Relativity and Cosmology research. The lunar ranging experiment; assembling the team, background observations with balloons and COBE, construction for the 1969 Apollo flight. Studies in black-body radiation generated by Dicke, 1964; Wilkinson joins project with Peter Roll and James Peebles; building microwave apparatus, need for helium load, mapping the horn and measuring insertion loss. Robert Wilson and Arno Penzias at Bell Laboratories publish background radiation papers ahead of Dickeís team, other researchers on the topic: Ralph Alpher, Robert Herman, James Follin, George Gamow, Joseph Weber. The American Physical Society (APS) meeting in 1966, measurements from Dickeís group confirms Penzias/Wilson support of big bang theory. Further discussion on lunar ranging, measuring gravitation, the Nordvedt effect, optical measurements with Roger Dube and Bill Wickes, Steve Boughn and Peter Saulson on infrared background measurements. Paper with Mark Davis (1974). Formation of young galaxies, ongoing DIRBE (Diffuse Infrared Background Explorer) on COBE (Cosmic Background Exteriment) designed to measure background radiation over wide spectral range CO measurements at high galactic latitude looking for dust clouds. CCD photometry by Ed Loh, find scale anistropy on NRAO maser with Juan Uson, large scale anistropy measurements. Discovery of dipole anistropy by Paul Henry and later confirmations, disconfirmation of quadropole effect. Plans for future work: g-2 experiments, measuring g and creating new infrared detector, search for low mass stars. Effects of being a physicist on his family life, hobbies, upcoming second marriage.

Transcript

Harwit:

I saw from your vita and also from your biography in The American Men in Science that you were born in Hillsdale, Michigan. Perhaps you can say a few things about early recollections of your family, family life, anything that you remember that interests you.

Wilkinson:

Okay, well, my family was middle-class. My father and mother both worked, my mother as a school teacher.

Harwit:

What does she teach?

Wilkinson:

She taught junior high school math. My father, who I think never formally graduated from high school, was a projectionist in a movie theater, but was interested in electronics, and probably was the most influential on me early on in choosing a technical career. He was always tinkering around with tubes and electronics, and he was good at fixing things; he was always fixing the car or something like that. Then, of course, my mother was good with numbers. Those were really about the only influences that would have directed me toward a technical field. I grew up in a very industrial-type setting in Michigan. Most of my friendsí fathers worked in factories on production lines.

Harwit:

Where is Hillsdale?

Wilkinson:

Hillsdale is in central southern Michigan, a very small town, primarily a farming community town, and, in fact, most of my family were farmers up until my parents. My grandparents were both from farm families.

Harwit:

What is the closest, bigger town?

Wilkinson:

The closest bigger town would be Jackson which is about 35 miles away, and then Detroit to the east, Chicago to the west. Hillsdale is about half-way in between. Itís farther west, even than Ann Arbor, just about in the center of the state. Lansing is straight north.

Harwit:

I see. Okay, fine. Now, did you have brothers or sisters?

Wilkinson:

I had an older brother, four years older, who is now an engineer for TRW.

Harwit:

What sort of an engineer?

Wilkinson:

Aeronautical. He works on Minuteman missiles and things like that — a lot of classified work so I donít really know what he does. He went as far as a masters in aeronautical engineering and then went right to work for the aerospace industry; moved to Los Angeles.

Harwit:

Did your parents from early on want both of you to study?

Wilkinson:

Yes, there was never any doubt about that. We heard about college from the time we were old enough to understand what it meant — a lot of emphasis, not pressure, but emphasis, encouragement to get good grades in school. We both transferred from the small town school into a larger school system, so we had to ride a bus.

Harwit:

How far away?

Wilkinson:

It was about 30 miles, so it was a considerable nuisance to have to ride this bus every day to school and back.

Harwit:

Where was the school located?

Wilkinson:

That was in Jackson, Michigan. I actually during the War moved to Texas, and after the War when we came back, we didnít go back to Hillsdale at that point. We were living in a little town named Michigan Center; itís a suburb of Jackson.

Harwit:

What made your family move to Texas?

Wilkinson:

My father was in the Navy. He was an instructor, electronics instructor in the Navy.

Harwit:

And did he go back to his old job when he came back.

Wilkinson:

Yes, after the War, they both went back to their old employment.

Harwit:

But they decided that they wanted to move closer to Jackson at that time.

Wilkinson:

At that point his job was in Jackson rather than Hillsdale. It was in the same local, the same area, same union but different town.

Harwit:

Did you and your brother interact much?

Wilkinson:

Not very much. The four years age difference meant that we didnít do that many things together.

Harwit:

I just thought maybe he would be building interesting things if he was interested in engineering, and that you would look over his shoulders or something like that.

Wilkinson:

Yes, he was always building model airplanes or...he built an 8-foot model of the Eiffel Tower out of toothpicks once and things like that.

Harwit:

Fantastic.

Wilkinson:

Yes, so I looked up to him as somebody who could build things and figure things out.

Harwit:

When you were in school, were you also interested in building things?

Wilkinson:

Not so much building things as tearing things apart and understanding how they worked. In that part of the world cars are a big thing, so every kid buys a car, probably even before he can drive it. Tears it apart, fixes the engine, that sort of thing. So I was interested in cars and mechanical things, not so much in electrical things.

Harwit:

Did you do that with other boys your age or by yourself?

Wilkinson:

Mostly by myself.

Harwit:

Did you go out and have a job in order to try to buy an old jalopy to fix up?

Wilkinson:

Yes, exactly.

Harwit:

At what?

Wilkinson:

I counted worms into a can — fishing worms. Fishing is big in southern Michigan, lots of lakes, so my job was put a hundred worms in a can, put the lid on. I donít know, we got a nickel a can or something for doing this.

Harwit:

You sold that to some vendor?

Wilkinson:

No, it was a worm hatchery. It was a guy who actually hatched out and grew earth wormsÖ

Harwit:

Oh, I see.

Wilkinson:

Öand I worked for him. But he needed somebody to package the worms.

Harwit:

So he actually got them to breed?

Wilkinson:

Oh yes.

Harwit:

And then you packaged then.

Wilkinson:

Then he would take them around and sell then to bait stores, locally.

Harwit:

How long can these things live after theyíre packaged?

Wilkinson:

Quite a while. You put some wet moss in there and they live for a couple of weeks.

Harwit:

Oh really, I see.

Wilkinson:

Just keep them moist.

Harwit:

A hundred per tin, then, or was it a box?

Wilkinson:

(laughter) It was like an ice cream container. So that was my first job to make any money. Thatís how I bought my first jalopy.

Harwit:

At a nickelÖ

Wilkinson:

Önickel a can.

Harwit:

It must have been a lot of cans.

Wilkinson:

But you could buy a jalopy for $50. Yes, a lot of worms.

Harwit:

Five thousand cans. Now when was it that you got interested in science or engineering? In school then?

Wilkinson:

Yes, really, I guess in high school. The school I went to, rode the bus to get to, was a very good school system. It had excellent math teachers; the physics course was a disaster, chemistry was good.

Harwit:

Did you realize that at the time?

Wilkinson:

Yes, it was quite obvious.

Harwit:

How did it feel? Was it that you hated physics?

Wilkinson:

No, not at all. I liked physics, but the teacher used to actually go to sleep in class. He was obviously bored with the whole thing. The course was taught in a very mundane way; it was boring.

Harwit:

When you say he would sleep, you mean that he actually fell asleep?

Wilkinson:

He would go to sleep. When we were doing labs, he would often just sit down at his desk and go to sleep.

Harwit:

Oh, I see.

Wilkinson:

He managed to stay awake during the lectures. (laughter) But it was a very boring, very mundane physics course; I canít remember much that I learned. But that did get me interested in science.

Harwit:

Not the physics, the chemistry did?

Wilkinson:

The chemistry and the math mostly.

Harwit:

Do you remember any teachersí names?

Wilkinson:

Oh yes. Strangely enough, the teacher I really remember was Miss Pitts. She was my English teacher. Miss Pitts had a way of getting you to think more clearly and to write things clearly than any of my other teachers, even my math teachers. She probably was more influential on me in high school than anyone else. She was a remarkable lady. A number of my high school friends whom Iíve met later, when asked that question, even though theyíre now engineers and doing technical things, still remember Miss Pitts as a very influential teacher. My math teachers, Mr. Holdeman I had for math and chemistry. He was a remarkable man, and very interested in what he was doing even though at that point; though he must have been at it for 30 years, he was still very excited about it.

Harwit:

So his enthusiasm transferred itself to the students.

Wilkinson:

And I had a geometry teacher whose name has slipped me at the moment, who was very good at stimulating us with the following mechanism: He would give a problem, and the first four people up to his desk with the right answer got an A, and the next four got a B and so on. (laughter)

Harwit:

So speed was essential.

Wilkinson:

Speed was essential — and accuracy. You had to have the right answer.

Harwit:

What happened if you came up early but didnít have the right answer?

Wilkinson:

You had to go back so youíd wasted time.

Harwit:

Oh, I see.

Wilkinson:

But he really put sort of a competitive spirit into your work, and I still remember that class because the brightest kids in my grade level were in that class. I guess that probably was the first time I was really able to sort out where I stood in math and technical things.

Harwit:

Were you always the best in your class?

Wilkinson:

Not by a long shot. In that particular class, occasionally Iíd be first, but usually I was third or fourth.

Harwit:

What happened to the others who were good?

Wilkinson:

Thatís a good question. One of them is a dentist in Jackson; the other one is a lawyer in Jackson; and the other one is an engineer in Los Angeles. He was the only one who ended up going into a technical field. The two brightest guys in my high school graduating class in mathematics and technical things ended up going back to the home town to be professional people.

Harwit:

The dentist and the lawyer?

Wilkinson:

Yes. They were both from well-to-do families, and numbers three and four, who were from middle-class families, working parents, both went into technical areas. I donít know what conclusion youíd draw from that.

Harwit:

Itís hard to say.

Wilkinson:

Maybe they just saw a lifestyle that we didnít see and thought they preferred that; but they were clearly better than the other two of us in mathematics, but decided very early that they wanted to do medicine and law — dentistry and law.

Harwit:

Were their parents either in the medical field or law?

Wilkinson:

One of them did have a...his father was a lawyer. The other one Iím not sure.

Harwit:

Was it the lawyer whose father was a lawyer also?

Wilkinson:

Yes. Iíve often wondered about that. What the influences were there.

Harwit:

Do you go back to Jackson sometimes?

Wilkinson:

Not very much. I still have an aunt living there, and I go to see her occasionally.

Harwit:

So you get a chanceÖor you have had a chance in past years to see some of these people you went to school with.

Wilkinson:

Yes, at least the ones that have stayed in the community.

Harwit:

How many from that school went on to college? What fraction?

Wilkinson:

For those days quite a high percentage — I would say over 50 percent, whichÖ

Harwit:

You started around 1953 or so?

Wilkinson:

Thatís correct. Graduated in Ď53 from high school. I think in those days 50 percent was rather a high percentage to go to high school, certainly the school I transferred from, perhaps the number was more like 10 percent.

Harwit:

So did some of them go to Michigan?

Wilkinson:

Yes, quite a few. It was the habit to go to nearby schools from that high school. A few came East or went West, but most people went to Michigan State, The University of Michigan or one of the colleges nearby — Albion — thereíre many colleges in that area.

Harwit:

Well, The University of Michigan was so good and the tuition for in-state people was low, so it must have been very attractive.

Wilkinson:

It was very attractive. In fact, I didnít even apply anywhere else.

Harwit:

When you got to Michigan, what was your first impression when youíd got there as a freshman? You had signed up for engineering, I guess?

Wilkinson:

Thatís correct. I got my degree in engineering. Well, my brother had gone there so I had had a bit of an introduction to the place. I wasnít overwhelmed by the size of the place. I guess my main impression was excitement. I was excited about learning, and felt I still had an awful lot to learn. And I worked hard as an undergraduate.

Harwit:

Was it very competitive?

Wilkinson:

Yes, yes.

Harwit:

More so than in a high school?

Wilkinson:

More so than in high school, I felt, at least, more stimulated by the competition.

Harwit:

Did you enjoy it, or did you feel bored?

Wilkinson:

No, not at all. I enjoyed it very much. I loved the hard work and the learning.

Harwit:

How about the social life?

Wilkinson:

Very little. I had very little social life. I went out of my way to get out of the dormitory. There was a rule that all freshmen were supposed to be in the dorm. I got a special dispensation to live out of the dorm.

Harwit:

Right as a freshman?

Wilkinson:

As a freshman. And that was partly on my brotherís advice who had lived in the dorm for his first year. He also worked hard as an undergraduate — realized that as soon as he moved out of the dorm, that that was more conducive to working.

Harwit:

Just quieter?

Wilkinson:

Yes.

Harwit:

So you lived in a room someplace in town in Ann Arbor?

Wilkinson:

Yes. So not much of a social life, true, but I loved it. Those are very happy years, looking back.

Harwit:

Was there time for dating also, or you said you worked pretty hard?

Wilkinson:

Yes, I tended to go back to Jackson quite often on weekends. It was only 40 miles away or so, and had a lot of good friends there. That really was my escape from the working. I worked hard during the week and then the weekends would relax.

Harwit:

So your friends were mainly back in Jackson not so much in Ann Arbor.

Wilkinson:

More back in Jackson, thatís correct.

Harwit:

And over the course of those years you didnít drift apart because your education was more advanced than the people who had stayed behind, or anything like that.

Wilkinson:

No we didnít drift apart, as a matter of fact. My best friend, Fred (now General) Nelson, from those years still, whom I see quite often, didnít go to a university. He went to the local junior college, but we stayed very close, did a lot of traveling together in the summer times, things like that.

Harwit:

What was the engineering physics program like at Michigan?

Wilkinson:

It was pretty much of a standard engineering program. You went through the same courses for the first two years. Toward the end you were allowed to take upper-class physics courses, rather than concentrating in electrical or mechanical, and so forth, I started out in electrical, but I came up against the steam tables.

Harwit:

(laughter) In electrical engineering?

Wilkinson:

Yes, there was a required course on steam engines, and the only book for the course was the steam tables, I decidedÖ

Harwit:

Keenan and Keens, or something like that?

Wilkinson:

Goodness, I canít even remember. I didnít take the course. Thatís when I switched into engineering physics.

Harwit:

I see.

Wilkinson:

But I was very happy I did.

Harwit:

You took math and physics courses mainly, then, the last couple of years?

Wilkinson:

Well, the math went all along because I was in engineering anyway.

Harwit:

Do you remember any professors in any of the subjects who influenced you one way or another?

Wilkinson:

Youíre talking about anywhere in undergraduate years?

Harwit:

Yes.

Wilkinson:

Oh, Dick Crane, no question.

Harwit:

I see. Well, you did your Ph.D. with him, didnít you?

Wilkinson:

Yes, but he was also my undergraduate E&M teacher, and thatís when I first saw a real physicist.

Harwit:

Whom did you have for math courses?

Wilkinson:

My freshman math teacher — I wish I could remember his name, because he was outstanding. I donít think I ever cracked open the calculus book because this guy put such beautiful lectures on the board that I just learned calculus from my notes. I donít ever remember opening a book. I wish I could remember his name because he...well you can learn calculus anywhere, but he really taught you to think clearly and reason and follow an argument, and he just made the whole thing so clear and understandable and logical.

Harwit:

Letís see, was that taught in the math department or in the engineering school?

Wilkinson:

That was in the engineering school. They had their own math faculty.

Harwit:

So all of the math you took was in the engineering school, is that right, even things like differential equations?

Wilkinson:

Yes, that was all in engineering; thatís right. So the math I took was more applied than I would have had if Iíd taken a straight physics degree, which I liked. I liked that side of it; I needed to be tied to real problems.

Harwit:

So you didnít have differential equations with Rainville, for example, who wrote a book and was in the math department at Michigan?

Wilkinson:

We used Rainvilleís book, as a matter of fact, but it was taught by an engineering faculty member. Churchill also was there at the time, I think, and we used his book.

Harwit:

That was on complex variables, wasnít it?

Wilkinson:

Yes, correct, gee, you know these books.

Harwit:

Now, you had a number of courses with the physics faculty, then, as an undergraduate,Ö

Wilkinson:

Yes.

Harwit:

Öso you probably had people like Laporte, Uhlenbeck, Dennison, at various times, either as an undergraduate or as a graduate student?

Wilkinson:

I had all three of them. As an undergraduate — Laporte for math-physics; Dennison for mechanics; Uhlenbeck, for statistical mechanics when I was a graduate student — all very good teachers, outstanding teachers.

Harwit:

Which of these influenced you most, or do you remember anything about them? Laporte was somewhat flamboyant and had mannerisms of all kinds, peculiarities, I guess.

Wilkinson:

Yes, Laporte was flamboyant. Well, strangely enough Laporte taught us how to keep a notebook. He, again, never used a book, used a text; he put the lectures on the board, and we were required to keep beautiful notes in ink with pictures and jokes. When we would sit and take our hour exams, Laporte would go through our notebooks and grade them. It was an interesting technique because it did force you to write things down with care, and it really taught me the value of a notebook. And I still have that notebook; I still go back and look at it occasionally. That was an undergraduate math physics course. If I were ever to teach such a course, which I wouldnít pretend to be able to do very well, I would haul out this notebook.

Harwit:

Just very well organized.

Wilkinson:

It had a lot of stories in it and key ideas.

Harwit:

He wanted you to record the stories too?

Wilkinson:

Well, I would make little notes about the stories. He would often tell stories. I donít think he ever told us to write those down, but I usually would make a note.

Harwit:

What about Crane as a lecturer? He was much quieter than the other people.

Wilkinson:

Thatís correct. He certainly did not come in and write beautiful lectures on the board. We needed to read the book, and his style was to embellish what you were already supposed to know when you walked into the room. I found that very good. A lot of students complained because they thought he should be working out problems, deriving Maxwellís equations, and so forth. Instead, heís telling us about antennas or polarization properties of waves and trying to stimulate us to think about more applied things that had to do with the material we were learning, and I enjoyed that.

Harwit:

Was Barker still teaching?

Wilkinson:

Barker was inactive; he was there, but at this point a very old man.

Harwit:

He had been chairman not too long before but might have been retired by the time you got there.

Wilkinson:

He was retired by the time I got there.

Harwit:

Dennison was chairman?

Wilkinson:

Dennison was chairman; thatís right.

Harwit:

Was Glaser still around?

Wilkinson:

Glaser was still there when I was an undergraduate, yes.

Harwit:

Anybody else there whom you remember as being influential on you in any way?

Wilkinson:

Oh gee, I think probably to some extent everyone in the department. It was an excellent teaching department at that time, and there were young people there who took teaching very seriously — this is while I was an undergraduate and a graduate student. Oh, people like Bob Lewis and Bill Ford, and they justÖ

Harwit:

These were younger people?

Wilkinson:

These are younger people who clearly spent a lot of their time on their teaching, and were always happy to talk to you, but were very up to date on what was going on. I learned an awful lot.

Harwit:

Hazen was also in the department, I think.

Wilkinson:

Yes, Hazen was there. In fact, I never took a course from him, but he did research on the same floor, and I would often go down and talk to him. He had some cloud chambers at that time.

Harwit:

So, when you decided to go to graduate school, what decided you to stay in Michigan? Was it the fact that Crane had interested you and you wanted to work with him ultimately?

Wilkinson:

Yes. That really was it. It was partly financial.

Harwit:

But at that point you were able able to get assistantships, so you could have gone elsewhere too, I imagine.

Wilkinson:

I could have gone elsewhere. I already had a job as an undergraduate with Crane, had been working for him for a couple of years doing various things; had worked a summer with Terwilliger and Jones on an electron accelerator — had enjoyed that. I think Dick Crane had even mentioned the g-2 work to me while I was still a senior, because I think when I walked in as a grad student I started out on the g-2 project from the very beginning. (Here g is the electron gyromagnetic ratio.) Yes, Iím pretty sure I did; I was working on that while I was still doing my coursework.

Harwit:

Now, Crane always was known to do quite difficult experiments which were very ingenious — seemed to be his style.

Wilkinson:

Yes, sure was.

Harwit:

Is that what attracted you to him, or was it a matter of temperament more? You know, when one chooses someone to work with?

Wilkinson:

I think a little of each probably. I certainly was very attracted by the physics that he was doing. I always found his g-2 idea very elegant, and a piece of fundamental work that could be done with pretty low-key techniques and without a lot of big machinery and technical support around. And that appealed to me, but Dick, himself, was a very attractive guy to work for. He gave you a lot of independence, but he was always there when you needed some backup. Inevitably, his advice was good and to the point and very helpful. I think I must have learned that as an undergraduate and saw that that was going to be aÖ

Harwit:

Temperamentally a good match then.

Wilkinson:

Oh yes, yes, a very quiet, unassuming man; but just remarkable insight and intuition into complicated physics problems — well not physics problems — just problems you were having in the lab. Something would look hopelessly complicated or involved — lots of different things pointing in different directions. Typically, heíd come up at four oíclock for a cup of coffee. Weíd sit down and talk. When I was having troubles, of course I would pour it all out. Dick would just sit there, take in all in, and be very quiet, and sometimes minutes would go by and nothing would be said. But Dick would be thinking about it, and usually heíd have something right away to suggest, but as often as not, heíd come back the next day with some advice.

Harwit:

How many graduate students were in the group that he had?

Wilkinson:

Me,

Harwit:

Just you.

Wilkinson:

Just one.

Harwit:

Was it his custom to have just one?

Wilkinson:

At that time, yes. He had his own carbon fourteen, C14, work which he ran himself. He had no students on that. At that time the only other research that he had going was this g-2, and through most of it I was his only student. Later on, Art Rich joined us as I was finishing up, and then for a long time Art was the only student.

Harwit:

So that was a contrast to, say, Laporteís shock-tube troops.

Wilkinson:

Yes, thatís right, a sharp contrast to that or to the cyclotron operation and so forth. And curiously enough Dick had come from the background of the synchrotron there at Michigan, which was big-time physics. It must have been a conscious decision on his part to go back to precision, smaller-time, almost atomic physics type things.

Harwit:

One has the impression with Crane that he was a much more complex person than people who perhaps were louder in the department.

Wilkinson:

Yes, yes, no doubt about it. Dick, very quiet unassuming guy, wasnít splashy, didnít teach field theory. I know many of the graduate students certainly didnít appreciate how deep this guy was, and he made no pretenses about knowing a lot of quantum theory or anything like that. But he had this remarkable intuition.

Harwit:

He did some beautiful experiments.

Wilkinson:

Oh yes, and he could work out things that were theoretical in nature. A good example is the whole idea of how this trap works in the g-2 [experiment]. Dick was getting results and had interpreted the data. He and Schupp did. Art Schupp was my predecessor. He was the grad student just before me, and he did the first trapping experiment to do g-2. All that work was done, interpreted, and published without ever a quantum theory of that trap or even a very good relativistic theory of the trap. That was mostly just conjured up out of classical physics, probably a lot of it from Dick, although I think Bob Pidd probably contributed some to that as well. Only later did Bill Ford come along with a real theory of that trap, a quantumÖwell, a relativistic, at least, theory of how that trap worked. Ken Case had analyzed the uniform field situation earlier on and showed that the classical models were really right even in a quantum theory. But Dick had intuited all this from the beginning.

Harwit:

Was Case at Michigan the whole time you were there, or had he already gone to the Rockefeller University by the time you finished?

Wilkinson:

By the time I finished, he had gone. Iím not sure he had terminated his appointment at Michigan at that point. He might have still been on leave and so forth.

Harwit:

What about Luttinger? Had he gone to Columbia already?

Wilkinson:

Yes. He had already left in the middle of my time there.

Harwit:

Now, when you worked on this problem for your thesis, you and Crane would talk and then youíd go off and do things on your own. You got through remarkably quickly; you got through in about five years altogether.

Wilkinson:

Partly because I started my first year, I really started on my thesis problem when I walked in as a first-year student. We emptied the laboratory, painted it even, and started all over again; so by the time I was a third-year student, when many people were just starting, I was already well along.

Harwit:

So it must have been a big advantage to have joined the group as an undergraduate.

Wilkinson:

Yes.

Harwit:

So you worked with Crane over a period of six, seven years.

Wilkinson:

Yes.

Harwit:

Heís long retired by now, isnít he?

Wilkinson:

Yes.

Harwit:

I forgot to ask you, was the bachelor in engineering a five-year program at the time?

Wilkinson:

No, that was a four.

Harwit:

A lot of schools still had five-year programs. I think Cornell only went off it, maybe in the mid-í60s, late Ď60s.

Wilkinson:

Michigan had a five-year program. You could sort of get a double major in, say, EE and physics, if you wanted to, or in math and EE and things like that. So it was possible to stay for a fifth year. I didnít see much point since I knew at that point I wanted to doÖWell, I didnít know I wanted to do physics, as a matter of fact, because I got a masters degree in nuclear engineering. I essentially spent a fifth year studying reactor physics, and then decided I didnít want to do that.

Harwit:

So then you came from Michigan straight to Princeton.

Wilkinson:

Yes. Well, I spent a year as a lecturer at Michigan after I got my degree. Stayed on a year, helped Art Rich get started for the next generation of g-2.

Harwit:

Did you have much interaction with the molecular people at Michigan?

Wilkinson:

No, unfortunately not. In fact, the molecular program was beginning to gear down at that point. Dennison was chairman, so he didnít have as much time to be active. There was no molecular spectroscopy going on at the time. Who was it still there doing theory?

Harwit:

Ted Hecht?

Wilkinson:

Hecht, yes. He later started doing nuclear physics, but he was thereÖHe was another of the young excellent teachers in that department at the time. His lectures ware just beautiful.

Harwit:

Okay, so, you stayed on then as a postdoc for a year, and what decided you to come to Princeton?

Wilkinson:

Well, letís see how did this go? The g-2 result was a good one for the times. It had surpassed Schuppís number by, I donít know, one or two orders of magnitude in accuracy; I canít remember. But it was the best g-2 number around, and in those days everyone was very interested in testing quantum electrodynamics, so there was a lot of interest in it. Peter Franken was on my committee, and, in fact, I had talked to Peter a lot while I was a grad student there. He was another very influential guy, more on experimental techniques and style, but I admired very much Frankenís originality in the kinds of experiments he did. Peter got on the phone and called up a bunch of people Ė DickeÖ

Harwit:

Peter Franken rather than Dick Crane?

Wilkinson:

Rather than Dick, yes, and thatís characteristic of their style. Peter is more flamboyant and would do that and Dick would sort of sit back and wait for somebody to call him or see somebody at a meeting, whatever. Anyway, all of a sudden these invitations started pouring in to go and talk at various places; so I made several trips to the East Coast. I think Peter called Vernon Hughes at Yale, and he must have called Ed Purcell or Norman Ramsey, probably Ramsey at Harvard.

Harwit:

Would it have been more in the line of Norman Ramseyís interest?

Wilkinson:

Yes, I think so.

Harwit:

But Ed Purcell also would have been possible.

Wilkinson:

So I made a couple of Eastern swings, and of course this was great, very exciting for me, mid-Western kid finally coming to the great Ivy League schools and giving talks.

Harwit:

You had never been on the East Coast before that?

Wilkinson:

No, I hadnít. No, in fact, my first trip was to Columbia, so Peter must have called I. I. Rabi, I suppose. Anyway, these letters started appearing or the phone started ringing, and that was pretty gratifying to get a phone call from people I had read about all these years and never seen. As a grad student I didnít go to any conferences.

Harwit:

There wasnít that much money in those days, I guess.

Wilkinson:

There wasnít that much money to travel around, and it was Dickís style. He did not go to a lot of meetings himself, and I pretty much followed his example. So I looked at various places and decided that I wanted to come to Princeton and work with Dicke on gravitation. I had decided I wanted to change fields. The g-2 is a kind of peculiar experiment; it wasnít really in a field; it wasnít a particle experiment; it wasnít atomic physics. It was off by itself, but I liked Dickeís style. He was doing the Eotvos experiment at that point.

Harwit:

Well, he also was doing interesting experiments, also sort of a one-man group, a few people, and also quiet but incisive?

Wilkinson:

Yes. I suppose that style was familiar and attractive too. Tom Carver had a good deal to do with getting me to come here. He was doing atomic physics, resonance, here. He, I suppose, originally came to work with or near Bob Dicke, because Dicke had some ideas about optical pumping and resonance experiments. In fact, Bob had had some fundamental ideas on how one might measure g-2 with great precision. So Tom was here doing optical pumping resonance, atomic-physics-type work, and he was quite interested in trying to get me to come to Princeton, I suppose thinking that I might work in his area as well. As it turned out, I never did.

Harwit:

You started out on lunar ranging. Is that right?

Wilkinson:

Thatís right. Thatís really the first thing I did here, at Bobís urging, and because of the times. It was pretty clearly a good idea. It wasnít original by any means, Bob and some of his students — I think they were all students, maybe not — had written a paper about the possibility of ranging to a reflecting satellite in orbit, and doing gravitation in this way. Then a paper came out with the idea of doing radio ranging by having a transponder on the moon. This would land a little transmitter-receiver, which would receive a pulse from the Earth, amplify it and send it back. Somehow we got a hold of that paper. I donít know whether Bob saw it orÖAs a matter of fact, I think Peter Franken saw it, and wrote to Bob or me and said, ďThis doesnít make much sense; you could do much better with a laser.Ē Franken at that time was bouncing laser beams off the surface of the moon, and knew about Dickeís earlier paper about satellites. I think maybe it was Peter who put this together and got that whole collaboration started. So I sat down and wrote a draft of the paper, andÖ

Harwit:

Of the proposal for such a thing?

Wilkinson:

Well, it was a peculiar strategy. What we decided to do was first write this short letter, hoping that NASA wouldnít go off on this lunar transponder thing but would rather think about putting up corner reflectors. We wanted to get something in the literature rather quickly and assembled this team of people who were interested in the project. I wrote a quick draft; we circulated it around. In fact, Iím not even sure if Franken ends up on the author list.

Harwit:

Yes, heís on it. I guess the authors are Alley, Bender, Dicke, Faller, Franken, Plotkin and you.

Wilkinson:

Right, okay. Mostly ex-associates of Bob Dickeís except for Plotkin, who was in NASA, and was interested in lunar ranging, and Peter Franken was also interested in lunar ranging.

Harwit:

And I guess you published that in the Journal of Geophysical Research, Vol. 70, page 2267, 1965.

Wilkinson:

Right. That was a project that I could work on, on the side without buying a lot of equipment, while I was trying to decide what I wanted to do.

Harwit:

Well, in any case, JGR, Journal of Geophysical Research, was the place where one would want to published something like that in those days.

Wilkinson:

Thatís correct. Also this paper about the transponder had been in there, so our interest really was to head off that idea if we could because a passive reflector just seemed like so much better an idea.

Harwit:

A lot of the early space literature, that is early in the NASA era, seems to have been published there, and I think thatís cooled down, itís become much more a journal for just what it says it is, rather than [space work of all kinds].

Wilkinson:

Yes, I think thatís right. Well, a lot of the early NASA science, of course, was the study of the Earth and the Moon and their interaction. Most of the conferences I went to in those early years, in fact, had to do with the Moon and the Earth and modes and tides, and all that sort of thing.

Harwit:

But it did attract people from a wide variety of different areas who perhaps had been drawn in by the fact that there was money available for research through NASA, that would support both that work and, maybe other things. Iím not quite sure.

Wilkinson:

Also it was pretty clear that a lot of the techniques that were going to become available were going to revolutionize measurements — getting satellites up there had a lot of advantages. Gravity, in particular; it looked like a great thing for gravity to get satellites in orbit.

Harwit:

Now, you started teaching right away when you came as an instructor?

Wilkinson:

Yes.

Harwit:

I guess you had an umbrella grant here through Dicke, or did you have to fund your own research with grants?

Wilkinson:

No, that was done with an umbrella grant that Bob Dicke had, and at that time there were several grants. He had some NSF money; he had some Naval Research money. Perhaps thatís all at that time. I think half of my academic-year salary and all of my summertime salary was paid by Bobís NSF grant, which he had, to do gravitational research.

Harwit:

So the university doesnít cover all of your academic-year salary here?

Wilkinson:

Not as a junior faculty member, and thatís still true. Senior faculty, yes.

Harwit:

What do you mean by junior in that case?

Wilkinson:

Non-tenured.

Harwit:

I see. So assistant professors are paid out of grants?

Wilkinson:

Half of their academic-year salary, yes.

Harwit:

By whom? Their own grants?

Wilkinson:

No, other peoplesí grants. The system is still very much the same: There is an umbrella grant in a certain area, say nuclear physics, and all of the young assistant professors and instructors in that group receive half of their academic-year salary from the grant.

Harwit:

But doesnít that make the assistant professors beholden to the people who pay them, and perhaps less adventurous than they might or should be at that age? Iím just asking. We donít have that system at Cornell.

Wilkinson:

I think that would depend very much on the head of the group. Whether or not when the new young person walks through the door he or she is told that youíre going to work on this project and do this. Perhaps that happens sometimes; I donít know. It doesnít happen in our group currently, and people are given a lot of freedom within the purview of the grant. If somebody in our group started doing nuclear physics, I think in a year or so we would have to make some adjustment, just because we wouldnít be ableÖ

Harwit:

Thatís what I was wondering really; the grants do require you to be true to the purposes that you stated within bounds, but nevertheless,

Wilkinson:

Fortunately, there still are some rather flexible umbrella grants around, and we feel very lucky, in fact, at still having one of these for our work here.

Harwit:

Itís not very often that you get something which is not quite specifically directed.

Wilkinson:

Thatís right, thatís right. Itís not very often, and weíre fortunate to have that. Weíve certainly taken advantage of it over the years. There are a number of things that would not have happened if weíd had to have written a proposal in advance saying exactly what we were going to do for the next three years.

Harwit:

Thatís right because it would have been turned down, I imagine.

Wilkinson:

It probably would have been turned down, and if when a new idea came along, we wouldnít have been able to work on it because we were going to have to write a report and justify our funding. Thatís an unfortunate trend, I think, in funding; people arenít given enough flexibility. I can see the things that drive the agencies to do that, but fortunately we still have an umbrella grant here whose title is ďGravitation, Relativity and Cosmology.Ē And thatís about as specific as the proposal gets.

Harwit:

And thatís a NSF grant?

Wilkinson:

Thatís a NSF grant. Thatís correct. Supports several senior faculty people here in research and half a dozen or so junior faculty.

Harwit:

And how specific do they expect you to be when you ask for renewals?

Wilkinson:

Not very specific at all. We say what weíre working on. First of all we say what weíve done. Thatís really the basis for the renewal in our view. Is our work up to par? Do we deserve to get some more money? Then what weíre working on currently and the directions we see our research going. I think ideally thatís the way science should be funded.

Harwit:

Well, itís unfortunate that itís not always that what you have done [that counts]ÖI mean, when a group has been in existence for 20 years like yours has here, or more even, what the group has achieved, particularly in the most recent five years or so, seems probably the best indicator of what itís going to be able to achieve over the next five years.

Wilkinson:

Yes.

Harwit:

But I think it does take a fairly broad-minded contract officer to appreciate that and not, under proposal pressure, to worry about the topic of the promises.

Wilkinson:

Well, itís not just the officer, too, itís also your peers and referees, because if he gets negative referee reports, thereís no way he can end up funding you. So the referees have to understand that too. It does not make it easy to write a proposal, because these days most of us expect a proposal to be very specific and outline a task, almost, and a set of goals, specific goals, and so forth. Our proposals donít do that.

Harwit:

So how do you handle that when you write your proposals?

Wilkinson:

We say really what I said: We tell what directions weíre going in; what we think weíre going to do for the next year, which can be quite specific; say what is going to happen to that piece of apparatus thatís two-thirds built; and when we intend to fly it and the results we think weíre going to get with it. Beyond that, you can only say, ďIím still interested in isotropy. I see the next generation of detectors being such and such; weíre working on that, and this is our idea for how to improve those experiments.Ē And thatís about it, and sometimes those ideas are not very specific, simply because you havenít yet worked out all of the engineering details.

Harwit:

Now, whatís the reception that you get from the referees for this kind of thing?

Wilkinson:

Very mixed, very mixed. Another problem is that because our work in our group is diverse, our referees are diverse. A fair amount of work we do is in optical work, so some optical astronomers certainly review the work. Peebles does his work in theory and model fittings, so clearly some mathematical people are going to get involved. Astronomers that are interested in cosmology, astronomers that are interested in the infraredÖ

Harwit:

How many people are involved in this grant? Who are the figures in that?

Wilkinson:

Okay. Dicke, Peebles, Wilkinson (Iím sort of going in historical order), Groth, Loh, Boughn, and Kuhn.

Harwit:

Boughn is a faculty member now?

Wilkinson:

Yes. Has been for six years. Steve Boughn came here from Stanford. Did gravity wave work before.

Harwit:

But he started out as a postdoc with you?

Wilkinson:

Actually started as an instructor.

Harwit:

And heís tenured now?

Wilkinson:

No. In fact, I think this is his year of decision. Ed Loh has been working in optical things with CCDs.

Harwit:

You did a paper together, I guess.

Wilkinson:

Thatís correct. Jeff Kuhn whoís working with Bob Dicke on the Sun. Juan Uson whoís here working with Jim and me and everybody — very flexible guy.

Harwit:

And heís also on the faculty?

Wilkinson:

No, in fact, heís leaving. Heís currently, I guess, in a research position, but heís been teaching on and off.

Harwit:

Where is he going?

Wilkinson:

Heís going to NRAO. I think thatís the list of the junior faculty, and then thereíre a half dozen graduate students. Joe Taylor has his own funding, although he certainly is intellectually a member of our group.

Harwit:

So, youíre talking about a grant, then, which is something like a couple of hundred thousand dollars or so?

Wilkinson:

No, up to about five hundred thousand dollars a year. This is supporting work on the blackbody radiation which I am doing; Jimís work on large-scale structure of matter in the Universe; Bob Dickeís work on the Sun, Ed Lohís work on using the CCDs to look at very deep galaxies and trying to do cosmology with redshifts of galaxies; gravitational wave work that is going on with Boughn and Kuhn, calculating how gravitational waves might make the Sun ring and could you detect those modes and so forth — a wide variety of things. You can imagine the problem in writing and refereeing this proposal.

Harwit:

How often do you have to renew it?

Wilkinson:

That varies. At first it was every year, then we went to a five-year cycle.

Harwit:

Oh, thatís nice.

Wilkinson:

Well, it was nice, except that it was very hard to write a proposal for five years, and I, in fact, requested a three-year period which we worked on for quite a while. We currently have a five-year proposal in again at the urging of the NSF. They thought they would rather be back on a five-year cycle. I thought the three-year was better in terms of being able to write a harder proposal. Itís just very hard to anticipate five years ahead.

Harwit:

Five years is very difficult.

Wilkinson:

Itís a long time, in particularly if youíre doing the kinds of things weíre doing, where at any point in a year you might just go off in some other direction.

Harwit:

You have a group of imaginative people.

Wilkinson:

Well, hopefully. If you look at history, in fact, we donít go off every three years in another direction, but we like to think weíre ready to do that if an important problem comes along. And this kind of long-term funding is very beneficial in that respect. Another reason we at one point requested a three-year funding cycle was that some of us felt we were getting a little too complacent and we werenít being chided enough by negative referee comments. In five years you can get awfully comfortable, and that had happened for a period of time, I think a three-year cycle is more healthy.

Harwit:

When you did the five-year cycles, you just had to write a progress report each year, then, and maybe negotiate the funding?

Wilkinson:

Yes, but only then with our contract officer at the NSF. Those were not reviewed, so we were not being subject to criticism.

Harwit:

The funding was just determined essentially on a straight level, except that the NSFís fortunes were going up and down somewhat, maybe.

Wilkinson:

Yes. And that was another reason for not staying with five-year funding: If we saw an opportunity, but that was going to require another person, then within that five-year constraint, we couldnít expect to get that kind of increment. So we in some sense lost some flexibility to open up new areas or add another excellent person to the group.

Harwit:

Now, you also have some support through the cosmic background explorer experiment, where you are on the team?

Wilkinson:

Yes, thatís correct. In fact, now there altogether are three NASA contracts that are more or less related to the COBE.

Harwit:

Do they support your ballooning too?

Wilkinson:

Yes, and the justification is that — and I think it is a good one — the instruments that are flying on COBE, in fact, have all been developed first as balloon instruments by various groups; and it just makes a lot of sense to keep pushing the instrumentation as long as you can. And the data from balloons also help the mission. COBE will do better, but whatever we know at the time COBE flies will help us do a better job on COBE. There are reasons for keeping working on this; and various groups in the COBE team are still doing ground-based work.

Harwit:

You sound defensive about it. To me, it just sounds very natural that you should be doing these things to make sure that you have the best instrumentation at the time COBE flies.

Wilkinson:

I guess Iím sounding defensive because we always have to defend this to NASA. For example, NASA has a hard and fast rule against supporting any ground-based astronomy. Well, that rule has, in fact, been interpreted that you shouldnít even be doing balloon or aircraft work on NASA funds. Well, clearly thatís ridiculous. How are you ever going to develop were flown by us — you and everybody else — years ago. And it seems obvious to us, but it does not to NASA people.

Harwit:

No, I know, which is strange, after 25 years, or a little longer, of NASA existence. It doesnít seem to have changed their style of working all that much.

Wilkinson:

If anything, in that respect, flexibility of instrumentation being able to put the latest thing up, theyíve gotten worse.

Harwit:

Yes.

Wilkinson:

COBEÖWe tried a few years ago — just a couple of years ago — to change the design for the COBE radiometers, and it was like pulling teeth. We had to sacrifice an instrument in order to get the other three upgraded to even half way to where they should be.

Harwit:

Which one did you sacrifice?

Wilkinson:

We had four radiometers at one point. For the isotropy we sacrificed the lowest frequency one, in order to get the other three cooled to improve their performance.

Harwit:

I see. Now, to get back to the lunar reflector experiment, you built most of that up here, yourself, at Princeton? Or what was the nature of the collaboration?

Wilkinson:

No, in fact, we did very little here at Princeton. The collaboration was very much a team, a science team, which got together and set direction, policy and strategy. Then various members of the team did the work, and, in fact, the Princeton members did very little of the real work. Carroll Alley was the group leader; took most of the initiative. He took the initiative in getting NASA to let contracts to industry to build the reflectors, to put together ground stations to observe this thing, to get things organized so that the packages got built and on to the systems. So Carroll really deserves a lot of the credit for making the thing work. Peter Bender did a lot of the initial calculations to determine whether or not the accuracy was going to be good enough and what kinds of things we were going to be able to do. Jim Faller worried a lot about ground station configurations, and, in fact, invented and built one himself at Wesleyan at that time. Henry Plotkin was very actively working at Goddard with telescopes and lasers, ranging to satellites, so his experience was really valuable in that project. Later on, some people, Derral Mulholland from JPL joined the team. He was very instrumental in getting together an ephemeris and programs to analyze the data. Those are the principal people who worked. Bob Dicke and I supplied a few ideas, I suppose, and kibitzed; but we didnít actually do any building.

Harwit:

Where was the actual construction done then?

Wilkinson:

The corner reflector itself was built by Bendix under NASA contract. That had to be done very quickly because that opportunity arose only because the ALSAP package did not get ready for the initial Apollo flight. That was a general-purpose instrument.

Harwit:

What would it have included?

Wilkinson:

Oh, it the gravimeter; some kind of soil sampler; cosmic-ray detectors, I think. It was an array of several instruments that ware in a package, and it became obvious around October, before the following Apollo launch which was to happen in June, that that package was not going to make the flight. At that point NASA designed a lead weight to take its place, and through no small effort on Carrollís part, they became aware that they might be able put the corner reflector box in there instead. Several people on the team dropped everything for a period of a few weeks. We cornered the worldís market in quartz and all of the skilled people around the world who could polish corner cubes, and managed to get the package together. Bendix worked hard to get this package designed, doing an actual engineering design. There ware a lot of worries. There was a lot of hard designing that still had to be done, only nine months before launch.

Harwit:

How big was this cube then?

Wilkinson:

Oh gee, it was an array of cubes. Each cube was about two inches in diameter, and there ware a lot of these things. That first array must have been two feet square, and it was a fairly close-packed, rectangular matrix of these cubes.

Harwit:

Was the quartz really that scarce?

Wilkinson:

Yes, it was not easy to get. We were surprised. Well, we had to get some that was already sitting on a shelf, and it turned out that some company in Germany had it.

Harwit:

This is fused quartz?

Wilkinson:

It was fused quartz. One would have thought there would be plenty of quartz around. It had to be good quality because one doesnít want to degrade the light inside the cube corner. We needed a good thermal design, and not much was known about the lunar surface. There ware worries about dust. Did you put a dust cover on? There ware a lot of these kinds of decisions that now NASA makes years ahead of time, that had to be made in literally a few days, and get this thing built and delivered to the Cape in April.

Harwit:

This was April of which year?

Wilkinson:

Oh my gosh, when did the first Apollo flight happen?

Harwit:

The first one that landed was summer of Ď69.

Wilkinson:

Thatís about right.

Harwit:

And it was on the first one that landed?

Wilkinson:

Yes, the team had gone along for some years at a low visibility. NASA didnít see this project as getting on to Apollo until later on. In fact, Iím not sure we were even assigned a flight at that point, but when they suddenly started looking around for something more useful than lead to put on the first lander, Carroll Alley was in there politicking and telling people that they should put this on.

Harwit:

I see.

Wilkinson:

Remarkably, NASA went for it and the thing got built.

Harwit:

They were more adventurous then than now.

Wilkinson:

Yes, much more, much more flexible.

Harwit:

I think itís the sane people who are in there, and theyíve just been bitten a few times and become too cautious. Is that possible?

Wilkinson:

I donít think so. I think itís a different generation, a different breed.

Harwit:

But itís still the sane faces.

Wilkinson:

No, I donít think so, at least not the people I deal with.

Harwit:

No?

Wilkinson:

Theyíre really quite different. Certainly more cautious. Thereís no doubt about that, and rightly so. You can understand why theyíre more cautious, but in those days there certainly was a lot more flexibility.

Harwit:

Why do you say itís rightly so, while at the same time youíre regretting it?

Wilkinson:

I can understand the bind theyíre in. They canít have that many failures. Congress will just cut their funding.

Harwit:

But they didnít have all that many failures then, right?

Wilkinson:

Thatís right. There werenít that many failures. Iím still amazed that Apollo went as well as it did.

Harwit:

Well, a lot of things are way overdesigned. You remember the Pioneers we designed, I think, for a one-year lifetime or something like that. Some of them were around and working ten and eleven years later and still may be, if one doesnít turn then off.

Wilkinson:

Yes, thatís right. Things are overdesigned, by the standards you and I have for something we would be willing to call an instrument.

Harwit:

Now, this clearly was an experiment which, because it didnít involve your personal construction and also because it was a long-term project and because it had to be built by Bendix, wasnít the a sort of thing your were going to continue doing here without also doing something else, I imagine, in order to get tenure, if nothing else?

Wilkinson:

Oh no, this clearly was a very low-profile project here. We did not spend a lot of time on this.

Harwit:

So, at that point somewhere along the line you got involved with the blackbody radiation as your major interest, is that right?

Wilkinson:

Yes, thatís right.

Harwit:

So the historical sequence was that you started in on the corner cube; got that rolling and then shopped around.

Wilkinson:

Yes, thatís right. In fact, I was shopping around all the time this corner cube thing was going on. It was clear to me that that was not a style I was interested in. I did not want to get involved with a large project with a lot of people like that. It would have been very easy; there were a lot of interesting problems that you could work on, but that wasnít what I wanted.

Harwit:

So, how did you get involved in the blackbody background? I read what you had written up for the Jansky lectures, but it didnít tell much about the internal goings on here at Princeton.

Wilkinson:

I can remember the day and the minute that I got involved in blackbody radiation.

Harwit:

What was that?

Wilkinson:

We had a great institution in those days up in Palmer. It was room 122. This was a room in the basement in which in one part of it Allan Shenstone did spectroscopy of triply ionized nickel. He was chairman here for some time — a grand man, who literally worked until the day he died, analyzing his spectra. Also, in this room I had a desk, and Jim Peebles had a desk. Bruce Partridge had a desk; Mark Goldenberg had a desk. Several graduate students did their research, and there were benches along the walls. There was a blackboard on one of the center posts which was about four feet by three feet square. Bob Dicke walked in one day, and we just started chatting. Someone else was there at the time, and I canít remember who it was. It could have been Bruce; it could have been Mark Goldenberg. Bob said, ďWell, you know, Iíve been thinking about oscillating universes, and it must be that the Universe really got hot when in bounced.Ē He really liked bouncing universes. Thatís what he talked about most at that time. ďIn order to get rid of these heavy elements and give us nice, pure hydrogen, the temperature must have gotten up to around 10 degrees back in the big bang, and it seems to me that there ought to be some radiation around. Given all the expansion thatís gone on since then, this is probably in the microwave band.Ē He drew a curve on the board, and talked a little bit more about what might have gone on.

Harwit:

Curve of what against what?

Wilkinson:

A blackbody curve, intensity against wavelength and said this probably is microwaves. The whole thing was pretty casual, and Bob often had ideas. He would tell us about them, and we would think about it. Later on they would go away. So at that point nobody, I think, including Bob took it very seriously, but he, I remember, he pointed out that his Dicke radiometer would be a nice thing to do this, but thereíd be these problems because the radiation was isotropic. Youíd have to get yourself a good absolute load and measure the total flux, and that wasnít easy.

Harwit:

Okay, Dave, you were just talking about the first suggestion of the background radiation.

Wilkinson:

Yes, right, and Bob, in that first little conversation, pointed out what the phenomenon would be like and mentioned that it could be measured, but that wouldnít be easy because of this absolute flux character. There was plenty of sensitivity, no problem there. Of course, he knew about the radiometer and its capabilities. Well, Iíd never done any microwave work, so the instrumental challenge didnít attract me. I was pretty skeptical of the whole thing; it seemed based first of all an oscillating universe, a bouncing universe, which was pretty speculative. I didnít get very excited about it, but I do remember it very clearly and what he said, and this probably went on only for 20 minutes, and then we went on to a meeting or something.

Harwit:

Do you remember which year it was and which time of year?

Wilkinson:

That happened only a few weeks before I finally made a commitment to work on the project. It was only a matter of a few weeks. The other person may have been Peter Roll, because Roll also hung around in that room. Iím a little sketchyÖoh, you asked about the time. I would have to go look at a notebook to see what the date was, I guess, but it would have been the order of a few weeks to a couple of months before I actually started working on this.

Harwit:

And when did you start working, or you donítÖ?

Wilkinson:

I donít remember now; Iíd have to go look at a book to get a date.

Harwit:

It would be nice to know it, possibly, there are a number of things that we could talk about tomorrow for a few minutesÖI mean, it would be good to know did it take you months, or weeks, or years to prepare for this? [1]

Wilkinson:

I could go back and look at notebooks and try to figure that out. I know at the time he mentioned this, I was still working on a manuscript for a book on electrons. In fact, itís a chapter in a book that came out in a series, so I could in fact look and see when that article finally got published.

Harwit:

You donít have that on your list here, though, I think.

Wilkinson:

Oh, isnít that in there? Thatís strange. I spent quite a lot of time when I first got here working on this thing — Properties of the Free Electron was the name of the chapter. Thatís strange; itís not in my publication list. I spent a hell of a lot of time on this. Itís not in there. Strange, Iíve never noticed that.

Harwit:

What was it called?

Wilkinson:

The chapter was called, ďProperties of the Free Electron,Ē and I went into all of the techniques for measuring charge, mass, g-2 of the free electron.

Harwit:

Where was that published?

Wilkinson:

It was published in the series on ďExperimental PhysicsĒ thatís put out by Academic Press, I think it is — many volumes of this thing. Itís edited by Marton, as I recall. This was in a volume on, I think, atomic physics, and I believe Vernon Hughes had something to do with that volume. He may have been a co-editor. Anyway, I spent quite a lot of time when I first got here working on that manuscript part time, and I think I was still working on it when Bob mentioned this idea so that may be a clue. Anyway, the sequence of events after that is a little vague, but I know it was a period of time before I agreed to work on this experiment.

Harwit:

What went through your mind as you decided, or was Dicke bringing it up over and over again?

Wilkinson:

Not very many times. He saw early on that this thing should be done, or he decided early on. He got Peter Roll to commit himself to it, because Roll at that point was finished with the Eotvos experiment and was not a participant in the solar experiment which Dicke and Goldenberg had just gotten going — and Hill. So Roll and I were kind of looking around or thinking about things. I think I was looking for something a little more classical and something to do in the lab. I was interested in astronomy, but had no background, and certainly didnít know anything about microwaves. I was thinking of hanging weights on strings yet, and still grappling withÖ

Harwit:

What is this an illusion to, that weights on strings?

Wilkinson:

Eotvos experiment.

Harwit:

Okay, fine.

Wilkinson:

I think, in fact, I was thinking about an active/passive gravitational mass experiment at the time, thinking about how to do that. Anyway, I was much more oriented toward the lab, and I suppose wanting to draw on experience more than I could in this microwave experiment. But I obviously didnít go search the literature to see what was going on. Roll and Peebles and probably Bob Dicke were in fact reading the Bell Labs literature on their radiometer experiments, looking at atmospheric radiation and so forth, so I probably was helping a little with that.

Harwit:

Ohmís things and JakesísÖ

Wilkinson:

Hoggís, Jakesís, yes, all of those people. We were reading that literature, everything we could find. In fact, Iíve decided we did not see the Jakes article. It was only recently pointed out to me by Dave Hogg.

Harwit:

Itís never cited.

Wilkinson:

And when I went back and looked at that article, I donít believe we had seen it, because he mentions 2 degrees in thereÖ

Harwit:

Yes, yes.

Wilkinson:

Öand itís really quite specific. We certainly did not see it beforeÖ

Harwit:

Ohmís is also fairly specific.

Wilkinson:

And Ohmís also, thatís right. We were at that point much more interested in the atmosphere and what kind of problems we were going to run into in the atmosphere. Anyway, at some point along the line, I suppose I ran out of ideas or hopes of doing one of these lab experiments, and capitulated and told Bob in the hall of Palmer that I would work on this with Roll. I think maybe he had been bugging me a little bit. So then Roll and I started building this microwave apparatus, and we had periodic, I think once a week, brown bag lunches in Bobís office. There were a number of people in the group involved in those lunches, but they were all geared toward this background radiation problem. Jim would talk about calculations and Peter and I would talk about the apparatus. I can remember some very specific detailed instrumentation discussions before we ever had a piece of wave guide in the lab. We had no equipment. Bob had some old stuff lying around from the Second World War, but it was either the wrong band or so ancient that we couldnít use it. Tom Carver had a little bit of X-band stuff around, but it was in use; we could only go look at it. So we werenít able to borrow anything; we had to go buy everything, and that took a long time, just to get stuff in.

Harwit:

And you had the money for it?

Wilkinson:

We had the money for it. It must have come from these umbrella grants that Bob had.

Harwit:

Money was easier in those days anyway, I suppose.

Wilkinson:

But I do remember Bob, very pointedly, saying we had to work at X-band, because thatís where the cheap equipment was available. There was no way we could work in K-band. We were to use Second World War surplus stuff where we could, and we went to Philadelphia and raided Arch Street in order to get pieces of wave guide.

Harwit:

Arch Street?

Wilkinson:

Arch Street is the war surplus electronics street in Philadelphia. But we did finally talk Bob into buying a few good things. We managed to get a calibrated attenuator — brand new — from Hewlett Packard and I think a klystron plus power supply for our local oscillator. Of course, we had to buy mixers and IF amplifiers. Those really had to be bought new. Most of the rest of the stuff we built. We built a precision attenuator, which in the end we never used, but we spent a lot of time on that thing. We built the horn. We built the cold load. We built the horn by Bobís specs. In fact, that was my first job was to build the horn antenna, and get out and measure it to make sure it had low side load response. Peter concentrated on the helium load, and we worked pretty independently on those two projects, realizing those were the key points. We needed good isolation. We needed to be able to tip the thing to take out the atmosphere. We needed good isolation from the ground, and we needed low reflection off both of these devices.

Harwit:

By isolation, you mean the back lobe had to be reduced.

Wilkinson:

Yes. Thereís this 300-degree ground.

Harwit:

Let me ask you. There are two conflicting things, or at least in my view they seem conflicting. One is a thing you once told me and another is one you wrote. I once asked you whether you thought one could have measured the microwave background radiation with Bob Dickeís wartime equipment, and you said absolutely no, you needed helium cooling. But in the article you and Jim Peebles wrote for the Jansky Memorial Symposium, you wrote that you thought one could have taken the instrument to a mountaintop and detected the radiation. Now, is there a conflict?

Wilkinson:

Yes, if I said that in that Jansky lecture, then I misled you, because you really needed that helium load.

Harwit:

Okay, yes, maybe I can point that out just since weíre here. Maybe I misread it; I was just looking at that again.

Wilkinson:

The only point of the mountaintop would have been to escape the water vapor radiation because that instrument was right on the waterline, but I canít imagine how you would get your absolute calibration — get your zero-point — without that load. You could do it the wayÖ

Harwit:

Yes, I didnít see it either; thatís why I was a little surprised, but I may have just misread. Okay, I think Iíve got it coming up here.

Wilkinson:

Iím thinking about how Jakes did it. Did Jakes have a helium load, or did he just add up every little piece?

Harwit:

I really donít recall at this moment.

Wilkinson:

He had a low-temperature receiver which helps a lot. Bob didnít; his receiver noise was horrendous.

Harwit:

I was guided towards that by John Pierce who had at one time been Director of Research at Bell later was at Caltech. He knew of my interest in tracing down the history of the discovery at one point, and so he forwarded a letter, or a copy of a letter, that, I think, Hogg had sent him, calling his attention to the Jakes thing, which I appreciated. I then went to the library, looked it up, and I have that reference in the Cosmic Discovery book then, because it struck me as interesting that the Penzias-Wilson observation was really the third one.

Wilkinson:

I guess thatís right. (laughter)

Harwit:

For some reason or another I donít seem to have theÖI just looked at it last night in fact. But anywayÖ

Wilkinson:

If I implied that in that Jansky thing, or said that, I think itís wrong; Iíll retract that, because even from a mountaintop, you still would not have a way of taking out things like the losses in the horn. You really need to either calculate very carefully, or have this absolute reference.

Harwit:

Yes, Iíve found it now. Now, actually you said that Bob had forgotten all about the measurement that he had done himself.

Wilkinson:

Yes, thatís right.

Harwit:

Who pointed that out to him?

Wilkinson:

Jim Peebles. In going through the literature about the atmosphere, Jim found a reference back to Bobís paper, went and read it, and this statement about the sky temperature canít be more than 20 degrees or something like that. Jim was the one who found it.

Harwit:

I see.

Wilkinson:

I remember we had quite a laugh about that. Jim pointed it out to him at one of these brown bag lunches. Bob was quite chagrined.

Harwit:

I thought that Jim was mainly doing the theoretical fireball calculations. He was also doing calculations on the other things — the atmospheric [factors], and so on?

Wilkinson:

I think he was also interested in the atmospheric problem, as I recall. He was reading the Bell Labs literature, along with Roll and me.

Harwit:

Iíll look this up afterwards, but at least I have the manuscript here.

Wilkinson:

We can talk about it tomorrow.

Harwit:

Now, you started building this then.

Wilkinson:

As soon as we could get the stuff.

Harwit:

And at what point did you then hear about the Bell Labs effort.

Wilkinson:

At what stage? We certainly werenít observing, and I think, in fact, I was still mapping the horn which took me a long time, to work out a technique to get down to the sensitivity I needed in the side lobes. First of all we had to build that horn, and Bob had some good ideas about how to do that. Then we had to set up and map it. Peter had this hard problem of building a reference source whose absolute temperature we were going to know. It was obvious the thing wanted to be down in helium which meant stainless steel wave guide, that was not easy to get a hold of in those days. It had to be plated; those things take time. It had to be measured. We had to know its insertion loss to get the radiation from the warm parts of that thing.

Harwit:

Itís what loss?

Wilkinson:

Insertion loss.

Harwit:

What is insertion loss?

Wilkinson:

Oh, you take a piece wave guide and put a watt in one end and you measure what comes out the other end; the insertion loss is what is absorbed. Thatís technical jargon.

Harwit:

Oh, I see, the matching, yes.

Wilkinson:

In fact, that points out one of the problems we had is that Peter and I knew nothing about any of this terminology. We couldnít have known a db fromÖAnd it took an enormous amount of time just to learn how to read the literature and talk to the manufacturers and specify the things we needed. Iím sure it took me an afternoon to figure out what a noise figure was in dbs, and we had to know that to even design theÖlet alone specify theÖinstrument.

Harwit:

Itís quite a different terminology from the optical one, even though one is talking about the same things.

Wilkinson:

It is, yes. Thatís right. So these two things — the horn and the cold load — needed a lot of very special measurement. We had to do things that were not even common in the microwave literature. You can there find out how to measure the loss in a piece of wave guide to 10%, which is usually good enough. We wanted to do that to a percent. Furthermore, we wanted to be able to do that over a whole range of temperature. That problem was not solved in the literature. Peter had to do that, figure out a way to do that, and do it from scratch. Of course, we had Bob Dicke to talk to; that was a great help. The temperature one was interesting. We needed to know the loss on the wall of that wave guide which had a bizarre surface, because first of all it was stainless, and then it had silver and then it had a gold flash; and you simply canít sit down and calculate because it depends on the surface texture and so forth, how well a plater did. Bob suggested making a cavity out of the thing and using the width of the absorption measure the Q. Most of that Q is due to the wall losses, and you use that. Then you can simply dunk this thing in liquid nitrogen and liquid helium, and you get the loss as a function of temperature that worked beautifully. In fact, I still get calls from people, occasionally, who read that long paper that Roll and I wrote about this experiment, asking about techniques that we talked about is one of them — this business of measuring wall losses as a function of temperature.

Harwit:

Thatís pretty fundamental.

Wilkinson:

Yes, Iím sure Bob Dicke had most of these ideas, but Roll and I, then, had to do it.

Harwit:

How did you and Roll divide up the work between yourselves?

Wilkinson:

Well, pretty much by parts. We decided which parts were going to be the hardest, and he took the cold load and I took the horn. Then we jointly worried about the specifications for the receiver and put the receiver together. Another tricky part was the Dicke switch, which Bob insisted be made symmetric — the one that was bought from the manufacturer — didnít have a plane of symmetry. Bob considered this a disaster, so we took it apart and put it back together with some extra wave guide around it and made it a symmetric switch. It was a very ingenuous idea and it worked.

Harwit:

Iím sorry, which of the papers with Roll did you mean, the one in Annals of Physics?

Wilkinson:

Oh, the thing in Annals of Physics which is a real tome. Its got all the experimental details in it. Apparently, people do read that thing; occasionally I get a phone call.

Harwit:

Well, itís useful to know things when theyíre done properly.

Wilkinson:

There were an awful lot of things that had to be measured with care, that just hadnít been done with that kind of care, except at Bell Labs. We read their papers with a great deal of interest because they had these cryogenic receivers, and so forth.

Harwit:

Did you go and talk with them also?

Wilkinson:

No, I donít believe so.

Harwit:

Öitís so close.

Wilkinson:

I donít believe we ever actually had a conversation.

Harwit:

This was the line that I was talking to you about, about the mountaintop. (Serendipitous Discoveries in Radio Astronomy, Ed. by K. Kellermann and B. Shetts, Natl. Radio Astronomy Observatory, Green Bank, 1984, page 176.)

Wilkinson:

Letís see what the context is here — Dicke radiometer invented by Dicke, used to measure the intensity; ďFig. 1 shows Dicke and his colleagues. This instrument on a mountaintop (to get above water vapor) could have detected the Fireball radiation, and, in fact, was usedĒÖokay. Alright, well, I could quibble, but youíre right, thatís very misleading — detecting it and measuring it, of course, are two different things. In fact, I think what led me to make this statement was that even in Florida, Bob was able to say that it wasnít bigger than 20 degrees by simply putting in all of the losses and things that he knew about without that cold load. Now the question is, suppose he had taken that to a mountaintop, which is what I was thinking about when I wrote that statement, to get rid of the atmospheric radiation. Could he have done it then? Thatís a question of how well he understood the loss in that horn, and all the losses down to the Dicke switch. If you can compute all of the losses on both arms of the Dicke switch with enough precision, then you can dispense with the cold load. It would have been hard.

Harwit:

Okay, I just wanted to know.

Wilkinson:

Itís a good point.

Harwit:

Itís been interesting. I talked to Dicke about that once. He thought he could do it.

Wilkinson:

He thought he could do it.

Harwit:

Ed Purcell thought he absolutely could not. (laughter)

Wilkinson:

I see, and Iíve told you both. Well, I must say, Iíve often thought of setting up an undergraduate lab experiment here to do it; and the thing that has always held me back, I couldnít figure out a way to do it easily without a helium load, and thatís more trouble than we like to put our undergraduates to. Maybe that puts, me back on the side of saying you really need that cold load.

Harwit:

Okay, itís sort of interesting because people have speculated. You see that in Weinbergís book, The First Three Minutes. He brings up the question of whether or not one could have done it with Dickeís equipment, and I think, if I remember correctly, he comes out on the side that, yes, people had told him one could have, and soÖ

Wilkinson:

My guess is, though, that you would end up in the same situation that Ohm, and Hogg, and Jakes ended up in. It would be so marginal, whether or not you could calculate and believe those insertion losses that you probably would write it offÖ

Harwit:

As they did.

Wilkinson:

Öas they did. If I were to retract, I would retract the statement in the Jansky lectures.

Harwit:

Well, you know, I wasnít trying to put you on the spot; I was just trying to understand what the situation was.

Wilkinson:

Thatís the situation. Those insertion losses, in getting from a horn down to a switch, minimum, are a few Kelvin, and that means that youíve got to do a 25%, maybe, calculation that you really believe, in order to get the uncertainty in those insertion losses below a degree. And there you are, 25% is tough because you donít even know the conductivity of the metal to a factor of 2.

Harwit:

Okay. Now, here you and Peter Roll were working on this, and at some point, I guess, you got a call from Penzias, was it, or fromÖ?

Wilkinson:

I think it was Arno Penzias, yes.

Harwit:

I know that Arno had been talking with Bernie Burke, and that Bernie had suggested that what you people were proposing theoretically might be what they were observing.

Wilkinson:

Right.

Harwit:

And do you remember what time of year that was, whether it was just before you started writing that joint set of letters that got sent of early in May of Ď65?

Wilkinson:

Yes, it was — my recollection would be — about a month before that, because we arranged for a visit. Incidentally, that call just happened to come in to Bobís office during one of these brown bag lunch affairs. The phone rang; we were all sitting there eating our lunch, and talking, I think, about an instrumental problem at the time. The phone rang, Bob picked it up, and we went on talking lowly, and then he said something over the phone about a helium load, and, of course, at that point we all perked up; and he started talking about the atmospheric radiation and reflections and so forth. So, it was quite clear that he was talking to somebody about this experiment, because everything he was saying were problems we were working on. I donít remember whether they arranged the visit at that point in that call, or whether they agreed to meet. I think he may have gone ahead and arranged the visit during that call, but Iíll never forget what he said when he hung up the phone. He hung up the phone, he turned around and said, in these exact words — I can hear him — ďBoys, weíve been scooped.Ē He knew immediately that they had seen it. He called us boys, which we were in those days. So Bob immediately realized the importance of that phone call. Now, I wish I could remember the exact stage we were at. Certainly we were measuring horns and cold loads. Whether we had the receiver working, I donít know. We certainly did not have the apparatus on the roof at that point.

Harwit:

It didnít take you very long to publish. I mean, you were publishing in Ď66, so you couldnít have been all that far behind.

Wilkinson:

When did the Letters come out?

Harwit:

The Letter came out, I think, In the July 1 issue. The date it was received was May 8, Ď65.

Wilkinson:

Okay, thatís makes sense. It was the spring when we went up to Bell Labs. Okay, yes, and it was that summer that we put the apparatus on the roof — summer of Ď65. So all of these precision measurements had been made before that. We understood the cold load, we knew the ground radiation from the horn, and I distinctly remember mapping that horn in the snow, and then on into the spring. So that activity was going on in the winter.

Harwit:

Who all went to visit them, then?

Wilkinson:

Roll and I, and Peebles and Bob Dicke, I think the four of us.

Harwit:

Peebles said he didnít go.

Wilkinson:

He didnít go? Amazing, I thought he was there. He would remember.

Harwit:

Well, this morning he said he didnít go, at least he doesnít remember it.

Wilkinson:

Okay. We went out in the cage and looked at the maser.

Harwit:

And you talked with Wilson and Penzias?

Wilkinson:

Asked them all the relevant questions about how theyíd measured insertion loss in all these various pieces of pipe? How they knew they werenít getting ground radiation — all these kinds of things into the horn. Even though before we went, we were quite sure that they were okay, because we went back and looked at the papers to make estimates of how big these problems would be, and in order to have missed that much, they would have to have had a pretty major problem.

Harwit:

But you hadnít had that feeling about Ohmís work, is that right?

Wilkinson:

Thatís right, and the main problem there was this business of not having that helium load. See, Arno had the great advantage that he could measure all of the insertion loss between the throat of the horn and the maser flange by sticking on this helium load, which he knew, which meant the only insertion loss he had to worry about was in the throat of the horn and on out. That horn was so big that heíd really have to have a lot of absorption on the walls, no metal could have emitted that much. So, it really looked likeÖ

Harwit:

So he had a small horn which he could cover with helium?

Wilkinson:

He didnít use a horn; he had a flange on the helium load, and he bolted it directly to the radiometer flange — the same flange that bolted the throat of the horn.

Harwit:

Okay, fine. So he could measure everything from the narrow end of the horn throughout his whole system.

Wilkinson:

And they taped all the cracks — you had to worry about cracks in the horn. They had thought very carefully about all these problems, and they had good, sharp crisp answers for everything we worried about. They had done careful work.

Harwit:

How much astrophysical thinking do you think they had done? Do you remember any of that?

Wilkinson:

Well, they had done an awful lot of thinking about the Galaxy, and what kinds of radiation could be coming in from the Galaxy, was my impression.

Harwit:

Well, that wouldnít have given you anything that was isotropic, presumably, right?

Wilkinson:

No, thatís right. Well, you could imagine a halo around the Galaxy.

Harwit:

Yes, at that time people were talking about galactic halos, particularly at Cambridge, Baldwin had been doing things like that. But I guess as they say in their paper, then, in a note that they append, they say that that wouldnít scale properly. Now this sounded more like an afterthought than anything else. I mean, the paper is remarkably empty of astrophysics.

Wilkinson:

Thatís right.

Harwit:

But they have this note added in proof where they say, ďThe highest frequency at which the background temperature of the sky had been measured previously was 404 Mc/s, where a minimum temperature of 16įK was observed. Combining this value with our result, we find that the average spectrum of the background radiation over this frequency range can be no steeper than λ0.7. This clearly eliminates the possibility that the radiation we observe is due to radio sources of types known to exist, since in this event, the spectrum would have to be very much steeper.Ē

Wilkinson:

Yes.

Harwit:

But elsewhere they donít have anythingÖItís almost as though they want to disassociate themselves from your explanation.

Wilkinson:

Well, they certainly were very skeptical of this whole idea, and it clearly came to them as a surprise or a new idea when Bernie told them about it. They, during this visit, did ask a number of things about the basic idea, the physical idea behind it.

Harwit:

See, the only astrophysical sentence they have is, ďA possible explanation for the observed excess noise temperature is the one given by Dicke, Peebles, Roll and Wilkinson in a companion letter in this issue.Ē

Wilkinson:

Well, that reflected very much their attitude during this visit. They were certainly very receptive to having this thing finally explained. I think they were sick and tired of the problem; they had had it with working on this and trying to figure out what the heck it was, and making all these detailed measurements.

Harwit:

Was it during this visit that you decided that you have these companion letters you would send in?

Wilkinson:

I donít remember. I donít remember when that was decided.

Harwit:

And you also donít remember whether it was in March or April, or when you would have beenÖ?

Wilkinson:

That we visited up there?

Harwit:

Yes.

Wilkinson:

I canít remember that either.

Harwit:

It wasnít too long before the papers were sent off, though?

Wilkinson:

I think the papers were written rather quickly. Jim and Bob wrote our paper. Peter and I essentially looked it over and made some comments, but I recall that happening quite quickly. My guess is it happened right after we came back from Bell Labs.

Harwit:

When did you become aware of the work that Alpher and Herman had done?

Wilkinson:

My recollection is that that happened in the winter after I made the measurements. The one clear thing I can remember is making the measurements up on the roof because it was fall and going into winter, and it was getting cold and I did it myself because Peter left Princeton that summer.

Harwit:

I see.

Wilkinson:

I was up nights and teaching the next day. It was a very tough winter and there was, of course, a lot of pressure because of this announcement. I think it was while our paper had been submitted, and I donít even remember who it was. Oh, I think I do now. Bob received a letter from Gamow at some point, and I think Bob came down and told me about the letter — kind of casually — and said, ďWell, I had one on him too because I sent a letter back pointing out something of mine that heíd forgotten.Ē And Iíve forgotten now what it was. It had something to do with cosmology, so it was kind of a tit for tat, jokingÖ

Harwit:

Gamow had done something which had not acknowledged Dicke?

Wilkinson:

It had not acknowledged Dicke, and Bob had pointed this out in the letter that he sent back. He made kind of a joke out of it. I think thatís the first I heard of it, so if you could get the date of that letter exchange, thatís when I heard of it for the first time.

Harwit:

Okay, so there is an exchange of letters between Gamow and Dicke in which Gamow complained and then Dicke sort of brushed it off.

Wilkinson:

Yes.

Harwit:

Alright, thatís good to know. Thatís good to know. But it was that way that you found out.

Wilkinson:

I think thatís how I first heard about it. Not a clear recollection, but I think thatís the way I heard about it.

Harwit:

You wouldnít, for example, have heard from George Field, who then remembered the McKellar and Adams CN work would not have known about the Alpher and Herman work?

Wilkinson:

I donít think so. In fact, I think if you go and look at Georgeís first paperÖ

Harwit:

I can ask him sometime.

Wilkinson:

Those early CN papers, I donít think they referenced Gamow, Alpher and Herman. I donít think so. That all sort of came up later. I think by the time of the APS meeting, where I got up and gave a talk about this measurement that would have been the APS meeting in the spring, I guess, of Ď66. By then, people were talking about Gamow, and I think in that paper, I mentioned and acknowledged the Gamow, Alpher and Herman work. In fact, I know I did.

Harwit:

Okay, so you gave a talk at the American Physical Society, then, in the spring of Ď66 on the measurements that you had made with the equipment you and Peter Roll had built up.

Wilkinson:

Yes, thatís correct. And at that time, that was the first number that seemed to agree with this idea and the Penzias/Wilson number. Otherwise, the big bang idea was really out on a limb, as here that was predicting a whole spectrum on the basis of one measurement.

Harwit:

Sure. Yes.

Wilkinson:

Nobody believed it. In fact, we got some really irate criticism about that Letter in the Ap.J.

Harwit:

Oh really?

Wilkinson:

Yes, oh yes. I had peopleÖ

Harwit:

From whom?

Wilkinson:

Oh, various conservative-type physicists and astronomers who said, ďYou guys must be nuts. Hereís this one measurement, and you guys are talking about radiation from the big bang.Ē They just thought that was utterÖ

Harwit:

It was only a Letter, after all, it wasnít as though you wereÖ

Wilkinson:

Well, I know; but the idea had in fact been circulating, because Jim had been talking about the calculation.

Harwit:

Iím surprised in his talks nobody would have mentioned.

Wilkinson:

Thatís amazing, isnít it, that nobody would have brought up the Gamow, Alpher and HermanÖ

Harwit:

I was saying this morning to him, I was surprised that Chandrasekhar would not have pointed this out when you submitted the Letter to The Astrophysical Journal, because he, himself, had at one time been involved. In 1942 he and Henrich had done a calculation of nucleogenesis in the early universe under static conditionsÖ

Wilkinson:

Yes.

Harwit:

Önot in an expanding [universe].

Wilkinson:

Yes, yes.

Harwit:

And so he must have been aware of the subsequent work of Gamow, of Alpher and of Herman, and theÖ

Wilkinson:

Thatís curious, isnít it?

Harwit:

Öinteresting thing is that Jim had, before you submitted this, Jim had, two months earlier, submitted a paper on galaxy formation in an initially hot universe which talked about this impending measurement of yours, to Ap. J., and nothing was said there. He also had sent in a paper to Physical Review the same day, or at least it was received the same day. Both of them were received March 8, 1965, which was turned down by the referees because they said it didnít contain anything new, and they referred to the paper by Alpher, Follin and Herman which you then also cite in your letter. Now, do you have any recollection of whether that was added right at the end, perhaps, when Peebles might have already received the rejection on that first paper? He couldnít recall this morning when I asked.

Wilkinson:

No, unfortunately, I remember very little of that letter.

Harwit:

Well, I think Jim would have had the rejection letter around the first of May from Pasternack, who at the time was the editor (of the Physical Review). But Pasternack evidently knew the right referee to send it to, and that referee referred to the Alpher/Follin/Herman paper because the paper that Peebles had sent in dealt with the nucleosynthesis.

Wilkinson:

Yes.

Harwit:

But the Alpher/Herman/Follin paper, which is quite remarkable because it predicts the ratio of neutrino to photon densities at the present time, did not give the 5į K, nor did it refer to that paper. So just having seen that paper, you would not automatically have been led back to either of the two (Alpher/Herman) papers, either the Nature paper, the short note in Nature they had written, or to the long Physical Review article they had written. Itís just one of those funny quirks.

Wilkinson:

Well, this business of the nucleosynthesis, my impression at the time was, that Jim had reinvented the idea, and then had come across this earlier work later. I didnít realize that it had been done in this way. Did that paper ever get published?

Harwit:

No, no it didnítÖ

Wilkinson:

Well, he eventually did publish a paper about nucleosynthesis.

Harwit:

Yes. Just the helium part.

Wilkinson:

I see. Thatís interesting.

Harwit:

That came out in Physical Review Letters in Ď66.

Wilkinson:

Okay. Did that refer to anything other than the (Alpher/Follin/ Herman paper)?

Harwit:

No, letís see, I have there here.

Wilkinson:

That didnít have the early radiation papers in it either?

Harwit:

No, I donít think so. Let me just check, I donít think he knew about it at the time. Okay, thatís called ďPrimeval Helium Abundance and the Primeval Fire BallĒ and it appeared in Vol. 16 of Physical Review Letters, page 410. He refers to the Alpher, Follin and Herman paper, but not to the original one thatÖHe may have not known at the time yet, unlessÖmaybe that hadnít been passed on by Bob Dicke, But the paper that he originally had sent to the Physical Review did refer to one of the Alpher/Herman papers which appeared the Annual Review of Nuclear Science, Vol. 2, in 1953. I went back and looked at that and it just has the original Alpher/Herman paper as one of the references on a tremendously long list. It doesnít single it out as having provided a specific prediction of a temperature for today. In fact, it doesnít mention the contemporary temperature. Itís almost as though the item of importance seemed to have been not the temperature but the chemical composition. Hayashi had written about that, and it seemed as though people were concentrating on the chemistry, getting the right abundances.

Wilkinson:

Yes.

Harwit:

There wasnít that much emphasis on the question of radiation, although Alpher and Herman had gone around and tried to get people to measure that early in the Ď50s.

Wilkinson:

Is that right?

Harwit:

Yes. They went down to the Naval Research Labs, which at the time was the only place where radio astronomy was being done in this country.

Wilkinson:

Gee, I would have thought any radio astronomer would haveÖ

Harwit:

I think they were scared by the fact that Bob Dicke had measured 200 with some difficulty, perhaps. Iím not sure.

Wilkinson:

Yes, okay. Well, also there was a lot of interesting, exciting radio astronomy to do in those days. They were probably busy measuring sources, and why spend two years doing this hard thing.

Harwit:

Itís still interesting, before we come off the topic of the blackbody radiation, which I donít want to talk to death, I was just trying to get an impression you might have for the fact that there is sort of a persistent[2]

Wilkinson:

That I would agree with, yes. They certainly did it, I think, carefully for the first time. Gamow never calculated anything in detail, as far as I can tell.

Harwit:

No, thatís true, yes. I would say in any case that itís certainly true that Gamow had the setting that he defined, and I would think it would be perfectly acceptable to attribute it to all three of them, perhaps.

Wilkinson:

Thatís generally what I do — attribute it to all three.

Harwit:

Itís always difficult to decide priorities, particularly where one person has been the professor of another person in the group. I mean, you run into this kind thing with Hewish and Bell, for example.

Wilkinson:

Yes, right.

Harwit:

Itís just very difficult for outsiders to know what went on blackboards that got erasedÖ

Wilkinson:

Sure.

Harwit:

Öconversations that obviously never got recorded and so on.

Wilkinson:

I think, though, in the end the history of physics really should be based on the literature, in at least priority disputes, because thereíre justÖwell, memories arenít good, especially when people get older, they fall more in love with their earlier discoveries, Öthey seem more and more important. In the end priority really should be established by the literature, I think.

Harwit:

Well, donít forget that the extrapolation you talk about here is not completely trivial. I mean, for example, the paper by Alpher and Herman worries about whether the microwave radiation would be seen in the presence of starlight density which also was known to be equivalent to 3į K in energy content.

Wilkinson:

Right.

Harwit:

Now, people have laughed at them for this, and maybe it was a blunder; I donít have any idea. On the other hand, nowadays when you know that about half of the starlight is absorbed by dust, and that dust could be emitting radiation at quite a long wavelength, for one knew at that time. Itís not such a stupid idea to say that youíd have to worry about that, although there might beÖ

Wilkinson:

Well, they didnít actually mention conversion or dust.

Harwit:

No, they didnít mention conversion, but Iím not sure whether it might not have been at the back of their minds. If it wasnít, itís again one of these rather remarkable blunders, because Bob Herman, when he was a graduate student at Princeton in pre-War days, had done some infrared measurements for his thesis work. You would expect a person like that to know what a blackbody spectrum is and that 5į K would be at quite a long wavelength.

Wilkinson:

Yes, thatís certainly is true.

Harwit:

So, you know, itís hard to know how that all fits together.

Wilkinson:

Yes, I donít know. On the face of it, the statement is ludicrous.

Harwit:

It is ludicrous, isnít it?

Wilkinson:

And itís hard to believe that two such good physicists would have made that statement, unless there were some other ideas behind it.

Harwit:

Well, they go through all this quantum mechanical general relativities stuff and then at the end of that they come out with a statement like that which just sounds very strange. But I guessÖ

Wilkinson:

Well, unless,ÖMaybe it sounds strange to us. Weíre experimenters and we immediately think of how you would go to detect this. Maybe they would think of putting a little black cube out in space and measure its temperature.

Harwit:

Well no, but I mean they did talk to the radio astronomers to try to see whether they could make the measurement, so they knew it was going to be in the microwave regionÖ

Wilkinson:

Right.

Harwit:

Öand they knew

Wilkinson:

But if you didnít know that microwave radiometers were up to this, itís likely they talked to a radio astronomer who threw up his hands and said, ďHow am I going to do this?Ē when my receivers are 5000į. So maybe they were thinking of putting a little cube out in space or making some kind of an energy, rather than a spectrum measurement. Then the idea isnít screwy at all, and the statement makes sense.

Harwit:

Yes, I donít know whether theyÖIíve never heard either of them say anything about a cube in space. I mean, theyíveÖNow of course this is with hindsight, becauseÖ

Wilkinson:

Yes, right. Now that you know how to do it.

Harwit:

But they say that they did go and talk to radio astronomers and that it just didnít seem to be something thatÖ

Wilkinson:

Let me tell you an a story that Joe Weber related to me only a couple of months ago thatís related to this. During the war, apparently he worked on radar, and after the war he wanted to get his Ph.D. He was at Maryland, apparently teaching, but hadnít gotten his degree yet, so he wanted to do a thesis and get a degree. Well, the most notable physicist around was Gamow. So, he went down to Washington, went into Gamowís office, and said, ďMy name is Joe Weber, and I have been doing radar for the war effort, and I know a lot about microwave electronics. Do you have any problems that I could work on, or do you have anything in mind that I could do?Ē Gamow practically threw him out — ďGo away, I donít want to talk about that, I donít know anything about microwaves.Ē

Harwit:

Interesting.

Wilkinson:

So Joe was offering his microwave expertise.

Harwit:

That was in Ď46 or so?

Wilkinson:

I wish I knew the date. Joe would know.

Harwit:

Okay, maybe I could ask him sometime.[3]

Wilkinson:

It would have been soon after the war, so I guess it would have the late Ď40s.

Harwit:

It might have been a little too early.

Wilkinson:

It might have been.

Harwit:

Okay, why donít we take a break at this point.

Wilkinson:

Okay, good.

Harwit:

Okay, we just took a coffee break for about 20 minutes. Let me start talking about some of the other papers youíve done, and since we talked about the lunar ranging earlier, why donít we talk about those for a while and then get back to background radiation and other problems. Maybe youíd want to start in on what you feel was the main thing that you got out of it. You did a number of different things in different years, and itís been a continually interesting experiment.

Wilkinson:

The lunar ranging.

Harwit:

The lunar ranging, yes.

Wilkinson:

Well, when the microwave background came along, I pretty much dropped out of the lunar ranging.

Harwit:

But you have papers with your name on them.

Wilkinson:

Thatís only because very early in the game there was an agreement among the people on the team that there would be a certain set of publications that would come out with all the team on. Thatís the only reason my name is on there.

Harwit:

I see.

Wilkinson:

Those papers are kind of an embarrassment to me. I probably couldnít tell you much about any one of them.

Harwit:

Okay, good.

Wilkinson:

The things that Bob and I did think about and worry about the most and, in fact, missed the main effect was the gravitational part of those. We were interested in what you might do with lunar ranging in order to measure gravitation, and then, of course, the Nordvedt effect was the Nordvedt effect, because he thought of it and realized you could get it out of these data. In fact, Bob had written down many times the Nordvedt effect, and neither of us noticed that. It was quite clear in the lunar ranging data that that effect was there and could be broken out and showed you something about gravitation. But our activity got to such a low level in that work, the later parts of that work, that we didnít spend any time.

Harwit:

Who wrote the papers then?

Wilkinson:

Oh my gosh, Pete Bender wrote a lot of them — the interesting data-analysis papers. Mulholland wrote a number of them because he was actually reducing data. Eric Silverberg probably wrote one or two of the instrumental papers. I donít know how many of those my name actually got on. I hope not very many.

Harwit:

Well, I think there are couple or three here that I was lookingÖ

Wilkinson:

The initial results, we had agreed, would be co-published.

Harwit:

You have one as late as Ď76 which I was just looking at recently.

Wilkinson:

Whatís the title of that thing?

Harwit:

Title is ďNew Test of the Equivalence Principle from Lunar Laser Ranging,Ē and that came out in Physical Review Letters, Vol. 36, 1976, (page 531).

Wilkinson:

Thatís the Nordvedt affect. I hope Nordvedt is on that paper.

Harwit:

No, you refer to him.

Wilkinson:

Okay, well I guess he predicted it in a paper.

Harwit:

Well, in that paper you put an upper limit to one of his parametrized post-Newtonian parameters, and you, of course, refer to his prediction and his classification of these parameters. Then itís interesting how tiny the errors are now that you accumulate over the years — I think it was 30 cm or something like that.

Wilkinson:

Now itís down to about 10 and headed for 3.

Harwit:

There hasnít been any degradation in the reflectivity.

Wilkinson:

As far as we can tell, no. The signals are as strong as they ever were.

Harwit:

So, the original worry about dust travelling on the moon electrostatically in some wayÖ

Wilkinson:

Didnít pan out. We were worried about meteorites hitting the area around and dust settling and all those kinds of things. No evidence that either the material has deteriorated. We were worried about whether the quartz would be fogged or pittedÖ

Harwit:

F-centers and things?

Wilkinson:

All kinds of things, yes — apparently not.

Harwit:

Wasnít there also a French/Russian device?

Wilkinson:

Yes. There was a thing that the French built and the Russians launched; it made a relatively hard landing on the moon. It had an eccentric weight so it was supposed to roll around and then end up pointing back toward the Earth somehow. That thing has been ranged to, as least itís reported that people have ranged to it; it gives a much smaller signal; and I think, in fact, weíre not using that because we have three of these arrays so we can already triangulate and get the lunar libration.

Harwit:

It wasnít a catís eye?

Wilkinson:

I donít think so. My recollection is that itís one big corner, a cube, that is intentionally degraded a bit. Youíve got to open up the return light in angle a little bit. We did it by diffraction — thatís why all these little corner cubes — otherwise, relativistic aberration moves the spot over and you donít see any signal.

Harwit:

I see, oh yes, thatís interesting.

Wilkinson:

That would have been an embarrassment to have missed that one.

Harwit:

Yes, that one would have been an embarrassment, wouldnít it. The moon moves fairly consistently, doesnít it?

Wilkinson:

Yes, you can imagine putting your receiver downstream a little bit. In fact, we thought about that at one time. Even with the built-in diffraction, if the receiver moved over a ways — I have forgotten how much it is now — you would get a stronger signal.

Harwit:

Okay, okay. All right. Well, why donít we look at some of the individual things you have done that donít quite fit into the larger pattern and then come back to the other ones. I was interested some of the optical measurements you made. I remember a few years ago you did this measurement that checked up on an original suggestion of Mattila in Finland, and you and Roger Dube did this, I think.

Wilkinson:

Yes, thatís right, and Bill Wickes.

Harwit:

Yes.

Wilkinson:

Right.

Harwit:

And, what got youÖ

Wilkinson:

Okay, actually, Mattilaís paper had nothing to do with that.

Harwit:

Is that right? I see.

Wilkinson:

The motivation was a paper by Partridge and Peebles, where they looked at galaxy formation and background light due to galaxies flashing up at high redshift and predicted the amount of general background light that might be left over from that. If the redshift isnít too high for that pulse of galaxy formation and if the stars are hot enough, some of that light could still be in the visible; so the idea was, go out and measure the extragalactic visible light and see if we can see anything. That was the motivation, Iím not quite sure when I heard about Mattilaís result. I think it came out while we were still doing this.

Harwit:

It could have. I guess, your first paper with the two others came out in 1977 and then you have a paper in Ď79.

Wilkinson:

Right.

Harwit:

Now, Mattila had published in either Ď75 and Ď76, I believe.

Wilkinson:

Okay. So we were aware of his result, at least at the time we wrote our paper.

Harwit:

Yes, I think you showed that your result was not compatible with the high extragalactic background that he was predicting.

Wilkinson:

Well, actually of all the experiments Iíve done at Princeton, that one probably I would consider the hardest and took the most nerve to try it.

Harwit:

Why is that?

Wilkinson:

Well, weíre lookingÖletís go to S10 unitsÖthis is a 10th magnitude star per square degree. Youíre not even in the ball game unless you can get down to 1S10, okay? If you look up at the sky at night, a dark night, you get 100S10.

Harwit:

Yes.

Wilkinson:

Thirty of that is from the atmosphere. Okay, you can handle that by doing tipping experiments like you do for the 3į radiation.

Harwit:

Sure.

Wilkinson:

Thirty of that is from starlight — individual stars in your field. Thatís harder, and 30 of that is from zodiacal light which just seems like itís almost impossible to think youíre going to measure zodiacal light to 3 percent. Certainly nobody even dreamed of if at the time we were thinking about this measurement. The literature on the brightness of the zodiacal light wasÖthere were conflicts of a factor of 2 about what the visible zodiacal light was at that time.

Harwit:

Is that true, I mean, Weinberg, I thought, had pretty good measurements.

Wilkinson:

I donít think his data from space were published at that point.

Harwit:

Not the data from space, youíre right, yes, I think youíre right.

Wilkinson:

We were still looking at data from Haleakala and some data from the Azores and so forth, and these people were disagreeing by factors of 2, at least.

Harwit:

I see.

Wilkinson:

It was not trivial to get all the numbers in the same set of units (laughter). Everybody used different units. Anyway, it took a lot of nerve to think we were going to go in and make a 1 percent measurement of all this, but the more we worried how to handle these systematic errorsÖ Oh, and the thing about the atmosphere that was really worrisome is that to 10S10 nobody could tell you what the fluctuations were like on a given night in any band. We were really going out on a limb. That experiment could easily have fallen on its face at the 10 or 20S10 level. In the end it worked out.

Harwit:

Worked out very nicely.

Wilkinson:

A lot of luck. The drifts overnightÖ

Harwit:

Why luck?

Wilkinson:

Well, it just happened that the half dozen nights or so we observed, we got two or three that were very stable, and two or three other nights that looked beautiful standing there the sky looked black and everything looked great. When you finally reduced the data and took out the atmospheric component, it was horrible. It was drifting by 10S10 over the night, and if weíd hit six nights like that, we would have thrown up our hands. The business of the zodiacal light, we had to make some checks on how well that — we looked in out of Fraunhofer lines — worked. There again, we didnít have a very good feeling in advance of how well that was going to work. When we did out checks on moonlight, evening, twilight and so forth, we were measuring the zodiacal light to 3 percent, which astounded us because it just hadnít been done from the ground with anything like that accuracy.

Harwit:

Yes, I think this in and out of the Fraunhofer lines was a very neat thing.

Wilkinson:

That worked beautifully, it always surprised me that nobody took that up that was interested in zodiacal light. Dube went to Kitt Peak with that mind. He was going to map the zodiacal looking in out of lines. So not being particularly interested in zodiacal light, I kind of dropped it and left it for Roger. Well, it never got done. Roger got interested in instrumentation projects.

Harwit:

Is he still at Kitt Peak?

Wilkinson:

No, he went on to JPL and then he went on to Michigan. Last I heard, he was back at Arizona.

Harwit:

I see. He had been an undergraduate at Cornell.

Wilkinson:

Thatís right.

Harwit:

But I think I just barely met him then one time. Hyron Spinrad did an experiment.

Wilkinson:

Yes. He redid Mattilaís experiment.

Harwit:

I guess he repeated Mattilaís and couldnít get that to work out either and did it more accurately.

Wilkinson:

Thatís right, and he says that his basic data disagree with Mattilaís. It isnít the matter of interpretation. The interpretation was shaky, and I always thought that thatís probably where Mattila went wrong, in worrying about what was scattered light and what was absorbed light off that cloud. Spinrad says no, that his basic observations disagree, so I donít know what went wrong.

Harwit:

The idea is certainly very nice.

Wilkinson:

Itís excellent and Iíve tried it myself in the infrared, using, in fact, the same cloud.

Harwit:

Sure, itís the best one around.

Wilkinson:

Unfortunately, itís not good enough in the infrared. That work never got published becauseÖ

Harwit:

What did you do there?

Wilkinson:

We went to Kitt Peak, and at two microns and one micron, looked on and off L134, because the second most important problem on my list is to find out what the infrared background is. Steve Boughn and I did this, and again the bloody cloud was brighter than the background. At this point we think itís backscattering of galactic infrared emission off the cloud.

Harwit:

Thatís the most likely explanation.

Wilkinson:

I would think so.

Harwit:

I wonder if one could again play the absorption band game with some infrared absorption?

Wilkinson:

We thought about using the so-called water absorption at 3.1 microns. Thatís such a narrow band that you lose in signal-to-noise too fast. We had to use a pretty broad band in order to get our signal-to-noise.

Harwit:

It would be interesting to have you publish that because, you know, a couple of years ago I suggested that to Steve Beckwith at Cornell, and he had a student who actually was looking at this a bit in a small telescope we have at Cornell — a 25-inch telescope. I mean, you know, you donít need a big telescope for that. They were just doing some preliminary observations, and decided against continuing it, butÖ

Wilkinson:

Thatís always the problem with not publishing negative or results you canít figure out. We didnít publish it because we never convinced ourselves that the scattering mechanism was the right one, and that there werenít some imbedded stars that we just didnít know about. We did apply for time to go back and take more resolution on the thing, even though we didnít think we were going to get the background, we were willing pursue it farther and see if we could understand why it didnít work. We were the only proposers that didnít get time that semester.

Harwit:

Oh, what a shame. I would think, I mean, you could discriminate against embedded stars, presumably, by having some sort of a chopper with narrow slits in it thatÖ

Wilkinson:

Well, it turned out that we had plenty of resolution. We were usingÖwell, I remember our beam compared to the cloud. It was about a tenth the size of the cloud, so we had plenty of resolution. In fact, the measurements we did get showed a large change over the cloud: It looked like the northern half of the cloud was bright and the southern half in fact was cold. So, we wanted to try and understand whether that was just a property of the cloudís surface or whether there were embedded stars in the north, but we needed more resolution to do that.

Harwit:

Itís such a fundamental measurement thatÖ

Wilkinson:

Steve hasnít, in fact, given up. In fact, I think heís going out to use the 1.3-meter next month.

Harwit:

Gee, it would be great. I mean, you know, to me thatís this whole galaxy background. I mean, itís been 17 years since Partridge and Peebles suggested this; and itís really a tough experiment, and yet itís the only cosmological test which is quite clearly defined and needed.

Wilkinson:

There are ways to wiggle out of it, unfortunately, if you donít see it.

Harwit:

All right, sure, but, I mean, it confines the era of star formation in galaxies somewhat.

Wilkinson:

Yes, there are lots of constraints you can put onÖ

Harwit:

The paper that you did on that point withÖin the red part of the spectrum on the —

Wilkinson:

Oh, with Mark Davis.

Harwit:

Right. Iím just trying to find it on your list. Yes, in 1974, Ap. J., 192, 251. That is also very nice.

Wilkinson:

That was fun. That was a lot of fun. I really stood there at the telescope expecting to see something. I thought we were going to get something out of that, because we were going a lot deeper than people had gone before. It just seemed reasonable that we might see something.

Harwit:

Now, that could be repeated at longer wavelengths, right?

Wilkinson:

Yes.

Harwit:

I mean, at that time, in 1974, one didnít yet have the really the ultrasensitive indium antimonide detectors we have today, and so you were able to set upper limits which referred to galaxies that would have been formed at redshifts less than, I think you said, 30, but then you would already have been at the blue end of the blackbody curve. Here you could go out with probably equal sensitivity to a factor of 3 more and look upÖ

Wilkinson:

Yes, that whole thing should be done as soon as thereís a good 2-micron CCD available from the ground that should be done again.

Harwit:

Are you going to work in that direction?

Wilkinson:

I havenít been thinking about it. Steve Boughn has, and, in fact, he has used the 4-meter in Chile and the 1-meter at Kitt Peak to do this kind of sky noise measurement with just beam chopping point at some blank sky and let it drift through and chop away and integrate to see if you see any roughness.

Harwit:

Yes.

Wilkinson:

So, the same kind ofÖ

Harwit:

Is he publishing that?

Wilkinson:

I think he has published that. He and Peter Saulson put out a little note. Itís a negative result but it does give a limit.

Harwit:

I guess I must have missed that paper because it certainly would have interested me otherwise, but I missed it.

Wilkinson:

Iíll have to ask SteveÖThis was Steveís idea, in fact, to use an infrared photometer and look for sky roughness. Itís similar to what Mark and I did, but I wasnít smart enough to think of doing it a two microns. It turns out his sensitivity can be quite good.

Harwit:

Okay, we were just talking about some experiments of Boughn and Saulson in the near infrared, and you were saying thatÖthat there are only upper limits, but you didnít say how constrained they are, whetherÖ

Wilkinson:

I think they end up putting constraints that are somewhat better than Mark Davis and I got in the red.

Harwit:

But going out to larger redshifts?

Wilkinson:

Yes.

Harwit:

So where are the young galaxies?

Wilkinson:

Theyíre probably around a redshift of 20 or 30, but theyíre just not as bright as we might have hoped.

Harwit:

Itís sort of interesting, the theorists now still are putting them at redshifts that are more like 5. If you put them at redshifts of 30, then what you were fearing at one point, in this paper with Davis, you point out one of the possibilities is the scattering by electrons. If you place the young galaxies that far back in time, it makes the electron scattering much easier, if you have any kind of ionized component in the intergalactic medium. And, in fact, if you have some residual gas which hasnít formed into galaxies at that epoch, in which caseÖ

Wilkinson:

Okay, wait a minute, Iím confused. This scattering from electronís was due to what?

Harwit:

Isotropize.

Wilkinson:

But thatís for the microwave background.

Harwit:

No, itís Thompson scattering. It can happen at all wavelengths, and I think you in fact mentioned that in your paper as one possibility forÖ

Wilkinson:

Ösmoothing it out so you donít see any roughness. Okay.

Harwit:

Thatís right.

Wilkinson:

At a redshift of 30 thatís hard to do. Youíd have to reionize the whole intergalactic medium.

Harwit:

Thatís right. But if you form young stars, the chances are youíve got enough ultraviolet light creeping out of the galaxies toÖ

Wilkinson:

Öreionize.

Harwit:

And then the density in the universe is still high enough. I think — Iím not quite sure — I donít know if it was Weyman or who, but I thought that the cutoff was around redshifts of 10. If you formed the galaxies earlier than 10, then the densities were sufficiently high so that if you ionize the residual intergalactic material, that you would get sufficient Thompson scattering. I donít remember, but it just sticks in my mind, I donít know where that comes from. I may be wrong on that. Itís just that itís one of those recollections.

Wilkinson:

Well, I think the thing Iíve worried more about over the years, whether youíre going to see them this way, is dust.

Harwit:

Intergalactic?

Wilkinson:

No, right in the galaxies. If you form enough dust quickly — dust really seems to form — it degrades the optical light.

Harwit:

The first generation youíd make?

Wilkinson:

Right.

Harwit:

So how would you do that? Youíd get a heavy material thrown outÖ

Wilkinson:

Övery quickly.

Harwit:

In a first-generation star? Are there any models for those? I always thought that the dust most people want the dust to come from a second generation. If you have dust in a first generation, then you could of course do it.

Wilkinson:

I donít see why you wouldnít get dust from the first generation.

Harwit:

It depends on how the material gets thrown out of the stars. IfÖ

Wilkinson:

I think these are going to be supernovae like mad, if thereís a whole bunch of 100,000į stars.

Harwit:

Thatís right. That depends then on the amount of the core that gets thrown out I suppose. Well, we know that the dust has to come out in some generation. It could come in the first, thatís true. So youíre saying the massive early stars could throw the stuff out and then block the radiation from the others.

Wilkinson:

Yes, dust is awfully effective.

Harwit:

And you would throw that into intergalactic space?

Wilkinson:

Yes, most of it.

Harwit:

Do you have hopes that COBE will tell you about young galaxies, just from the background?

Wilkinson:

Yes, I think the DIRBE experiment on COBE is the most exciting one on there. Thatís the infrared instrument.

Harwit:

Letís see, DIRBEÖwhat does the acronym stand for?

Wilkinson:

Diffuse Infrared Background Explorer.

Harwit:

Okay, and COBE is Cosmic Background Experiment.

Wilkinson:

Right. And as you know DIRBE, thanks to you, goes all the way from a micron to 300 microns.

Harwit:

So it should be pretty nice.

Wilkinson:

And itís an absolutely calibrated thing. Everything is cold; there is a cold reference load, so that thing will measure background radiation from all kinds of things over a wide spectral range.

Harwit:

That should really be an exciting.

Wilkinson:

Yes, it should really open things up in that band. There are a lot of foreground sources, but itís hard to believe that there isnít going to be something really interesting. The driver for that experiment was to look for the extragalactic infrared background. For that you need all the spectral-spacial coverage you can get to try and unsort the zodiacal component and all the rest of it — whatever scattering there might be.

Harwit:

It should really be great.

Wilkinson:

Yes, I think thatís going to be a very exciting experiment. The anisotropy will be interesting, but I donít think there is going to be anything fundamental coming out of that.

Harwit:

Do you expect that youíre going to be able to do orders of magnitude better than what youíve already been able to achieve?

Wilkinson:

No, not in sensitivity. In sky coverage it will be better because weíll cover more sky faster, which means we wonít have sort of systematic problems in trying to match one part of the sky in one set of observations onto another. Weíll have better frequency coverage. Itíll just be much more systematically taken data, and data that you can interpret and have more confidence in. In sensitivity, for most of those frequencies we wonít go much past what weíve already done in limited regions of the sky, but just because the instruments are old-fashioned by nowadaysí standards.

Harwit:

Let me, before I get into this, talk about one other experiment that youíve done that fascinated me, and I was wondering what decided you to undertake it and this was the relatively recent, line-of-sight CO measurements that you did at high galactic latitude? What led to that?

Wilkinson:

Thatís a long story but not unrelated to what weíve been talking about. Our experience with L134 in the infrared made us think that itís a bad candidate, and rather than beating on that and trying to understand all the physics if it is scattering, thatís going to be very complicated — probably hopeless. Why donít we go find a cloud out of the plane and further away from the center and work on that, because all you need is one example where the signal goes down. You look at that cloud, and youíve got the number and you donít have to worry about all this complicated interpretation. That was a way to bypass some hard work. We were looking for clouds high enough. The reasoning was this: most people tell us that hydrogen is associated with dust — that seems to be true from the Heiles maps — so letís go find some places where (a) we donít see any galaxies at high galactic latitudes and (b) there seems to be an anomalous amount of hydrogen. That might tell us that there is a cold, dark dust cloud sitting there that wouldnít have been seen. Okay, suppose we have one of these things — and the whole idea is to use it as screen for looking at infrared background — how would we know it? Well, we could go to Kitt Peak and ask for time again, and thatís even crazier. Weíre certainly going to get turned down if we promise to look at some unseen cloud of high galactic latitude. The idea was to look for the CO emission in this thing, and that was the whole idea, to find a high latitude cold, dark cloud, Rachel Dewey found a couple of spots in the sky to go and look at this. We got Tony Stark at Holmdel to help.

Harwit:

Yes, yes. Tony Stark, heís here.

Wilkinson:

He gave us some of his time on the Bell Labs telescope. We pointed at these two blank spots in the sky and looked for CO — integrated like mad, and it wasnít there. The current thinking on that is that the correlation between dust and hydrogen breaks down at low densities, and there even now is some physical evidence that thatís true. So probably if there is a dust cloud thereÖ Well, there is hydrogen; we know that from the Heiles map. Thatís how we picked that spot; it had a lot of hydrogen. What weíre thinking now is that that doesnít necessarily mean dust.

Harwit:

Well, I guess Heiles found that there were regions where he didnít get any 21-cm radiation, which were holes and indicated the presence of molecular hydrogenÖ

Wilkinson:

Yes.

Harwit:

Öand the CO tends to be associated with the molecular hydrogen more than with the atomic, I guess. At least, thatís my impression.

Wilkinson:

That can also happen probably because there is enough radiation around to break down the CO if it isnít protected by the dust, so you might expectÖ

Harwit:

Öwould dissociate the CO?

Wilkinson:

Thatís right. You might expect to be associated because of that. There were a couple of reasons to think that in fact this was not a good way to look for dust clouds, but weíre still very interested in that problem.

Harwit:

But it still might be a good candidate for you to attempt this extragalactic background [measurement].

Wilkinson:

Weíre red hot to find a good candidate at high galactic latitude, and weíre now working on the IRAS data to see if we can find it that way.

Harwit:

Just by looking for a cloud that would emit at 60 and 100 microns with no optical counterpart.

Wilkinson:

Right, and be resolved — barely resolved by IRAS.

Harwit:

And almost nothing in the shortest wavelengths on IRAS.

Wilkinson:

That would be a candidate. The other thing that weíve tried a couple of times now is that Ed Groth has some very deep plates taken with the 4-meter, and he has digitized these things and located all the galaxies on these plates. Heís now written a little statistical routine to go through and try to find any anomalously blank little spots which have a high statistical probability of not existing. Thatís a very promising technique, as a matter of fact.

Harwit:

Does he find such spots?

Wilkinson:

Yes, and weíre going to try and correlate those spots withÖ

Harwit:

Do you think youíll get turned down for time on Kitt Peak?

Wilkinson:

Not if we find a spot that we can make a case for. I think weíll get time. That was an anomalous situation. I think probably we had a classical infrared ground-based astronomer reading the proposal and saying, ďWe shouldnít be wasting our telescope on this crap.Ē (laughter)

Harwit:

Itís always the interesting things, isnít it.

Wilkinson:

Well, sometimes everything depends on who reads the proposal and whoís refereeing it. Sometimes theyíll go for the crazy things.

Harwit:

Sure, thatís true. What about the CCD photometry that you briefly got involved in and maybe still are, I donít know?

Wilkinson:

No, I gave that up to get back to isoptropy. Well, we got involved with that after the Mark Davis work, because obviously these things had better quantum efficiency, the spatial resolution, they were broadband, it was exactly the thing to go look for wiggles in the sky at very red wavelengths. Thatís why we got interested in developing that.

Harwit:

Let me identify both of these papers that we just talked about for the record, the one about the CO emission is in The Astronomical Journal Vol. 88, page 1832, 1983, and the CCD photometry of two distant clusters is in The Astrophysical Journal, Vol. 255, pages 57-64, 1982.

Wilkinson:

Yes, the CCD work really was motivated by wanting to look as deep in the red as you can and get some spatial resolution, and Ed Loh has pushed that problem to the hilt, and itís now paying off in spades. Within the next couple of years weíre going to be seeing some very fundamental things coming out of that work.

Harwit:

What do you expect youíll be primarily seeing?

Wilkinson:

He is now measuring redshifts of hundreds of galaxies at a time out to a redshift of 1 by an old technique invented by Baum, looking at broadband colors of galaxies. What that means is you can go out fainter and you can do a whole bunch of them at a time. This allows him to do cosmological tests, because now he gets the red shift as well as the magnitude.

Harwit:

No, letís see, he looks at colors first and then he goes to individual galaxies and measures redshifts?

Wilkinson:

No, he takes a CCD picture through five different filters, each a 1000 Å wide, and goes very deep so that he sees galaxies in each of these bands out to a redshift of one. He then analyzes those data — gets a magnitude in each of those color bands, he then compares those magnitudes for each individual galaxy to templates of all kinds of galaxies, all kinds of stars, at all different red shifts, so the computational problem here is enormous, but heís now managed to do this. He is getting redshifts and magnitudes for enormous numbers of galaxies.

Harwit:

The most promising ones he checks individually then?

Wilkinson:

No, at this point heís still doing statistical things. Heís pretty sure heís getting the redshifts right because heís looked at a few clusters and then does blind tests to make sure he sees that cluster in the redshift data. Heís already applied this to the following test, which is a little different from any of the other tests: He measures the density of galaxies as a function of redshift — a kind of a log Z-log S-type test, I guess — and looks to see if the density is changing with z. From this he can find out what Ω is; he can get the acceleration parameter, and from one run — heís using the Wyoming 90-inch infrared telescope to do this — heís already gotten quite a good number for Ω; he thinks itís one, within 20 percent.

Harwit:

That omega is one?

Wilkinson:

Yes, and heís now checking for all the systematic effects he can find. You can imagine with that much data, thatís an enormous job.

Harwit:

Yes, that is very interesting. I can see why you are so enthusiastic.

Wilkinson:

Yes, itís a fascinating technique. Iím very high on that work. Well, I made a deliberate decision about three or four years ago, not to go on in that. I spent quite a lot of time in the development of the detectors and in some early measurements about halos around galaxies to see if they were infrared halos, to see if the missing mass was in M-type stars and colder. Then, really just as the problem started getting interesting, and the real extragalactic work had started, which is what originally motivated it, I left it and went back to anisotropy.

Harwit:

So letís talk about anisotropy. You have done almost two decades worth of ever increasingÖ

Wilkinson:

I know; itís disgusting (laughs).

Harwit:

Not at all. You know, I mean, think itís really interesting stuff, and the only way that one can do it is by persisting as you have, you know.

Wilkinson:

I finally decided that was true. In fact, the reason I went off on this CD tangent was I was really hoping somebody else would come in and get involved with it, and that enough other people would get involved with it — that the problem would get resolved. And I was feeling a little stale. It didnít happen which was a big surprise to me.

Harwit:

Who would have come in?

Wilkinson:

Well, George Smoot and the Berkeley group and Phil Lubin have come in, and theyíre very good, but you need more than one group to be doing credible work.

Harwit:

Thatís true, yes. I think something like this needs two quite different types of instrumentation before people start believing anything.

Wilkinson:

Yes, thatís right. And the Italian group, Melchiorriís work has contributed significantly to large-scale anisotropy.

Harwit:

I notice in the most recent Astrophysical Journal, I think itís the August 15 issue, you have the lead article. There you are down to small-scale anisotropy of at level of a few times 10-40įK?

Wilkinson:

Weíre below that now. Weíre down to two parts in 105 at the moment. That sounds like quibbling, but in fact these factors are so hard won that itís getting so that we do quibble about whetherÖ

Harwit:

Yes, it was 5x10-5įK, I think, wasnít it?

Wilkinson:

That sounds about right, yes. Divide by 3 and you get, you knowÖOkay, thatís in fact a whole new line. Thatís a different game completely than the large-scale stuff from balloons.

Harwit:

Yes, yes. That was done with the 140-ft. at NRAO.

Wilkinson:

Yes. In fact, thatís an interesting story. Juan Uson came here a few years ago with a very formal mathematical background in Spanish universities — an impressive set of papers, but so formal-looking that I couldnít believe that he would even talk to Jim Peebles. I thought sure heíd end up talking to Arthur Wightman or Elliot Lieb or somebody upstairs.

Harwit:

Who is Arthur Wightman?

Wilkinson:

He is our ace math physicist. I thought even though Juan came to visit with our group, he would talk to Jim Peebles about some mathematics and then get interested in something else. Well, low and behold, he walked into my office and said he wanted to do an experiment. I shuddered and thought of something harmless like fine-scale anisotropy which had been going on for years, but I had noticed that the NRAO had a good maser. In fact, a twin of our maser that we do the large-scale anisotropy. We have a maser that we do the large-scale with, built at JPL. The NRAO, at Greenbank now, has its twin on the 140-ft., although much improved by the NRAO engineers. Just from a plain signal-to-noise point of view, it looked like, if you were lucky, you might be able to do a factor of 2 or 3 better with that instrument, and whatís the harm in applying for some time and trying it. So Juan applied for some time and sure enough they awarded it. Then the work began because from the literature we knew that we were going to have a lot of trouble from all ground radiation, atmospheric fluctuations, and all these really dirty problems with using big ground-based telescopes to do high sensitivity work. We went down there, and sure enough, we had all the problems, plus the maser didnít work very well either. It had a lot of noise in it, and we looked at the power spectrum at low frequencies where we wanted to beam switch and that was a mess. It was pretty discouraging; I was about ready to give up. Well, Juan just grabbed this problem, and he turns out to have a very good intuition for instrumentation and measurement. He has a very good instinct to go for the throat. He knows exactly where the problem is and heíll go right to it, plus heís very good with people. He started talking to their engineers and their telescope scientists and so forth, pointing out what these problems were; and sure enough, one by one these problems got cured. Weíd go back for another observing run and things would be a little better, and weíd complain about what was wrong — not complain but point out nicely what was wrong — and how much better the experiment could be. Meanwhile, we were also improving our scanning techniques. If you want to get to micro-Kelvins with that telescope, any telescope, youíve got to treat the ground radiation with extreme care. The scanning technique — how you scan the source in the sky — is very important and depends on the details of what all this complicated side-lobe structure looks like, what the horizon looks like, how the ground is changing temperature, and all these things that you canít predict. Youíve got to go to the place and make a number of mundane but important experiments that may take hours of integration to figure out whatís going on with all these systematics. Well, one thing led to another and miraculously these problems kept disappearing or at least going down below the statistical errors. Finally, after about four observing runs down there, we started getting numbers that were competitive with what other people had gotten. Still, several factors away from what the signal-to-noise calculations with just the maser would tell you, you could to. But Juan just stayed at this thing and was very clever inventing ways to beat the instrumental problems, and we[4] — thought carefully about how to scan and cancel out the ground. The result is this thing is now down to two parts in 105, which is really pushing the old theorists now.

Harwit:

Thatís right, yes.

Wilkinson:

Theyíre hurting, and it was this last factor of three or four that really started pushing them.

Harwit:

Do you think thatÖI guess itíll take some time to get down to that sort of thing in the large-scale anisotropy, right? Because right now one is dealing with a few parts in 10,000?

Wilkinson:

Well, it depends on what you mean. If youíre willing to take all of the large-scale data, average over the sky and say, ďWhat could be there in a quadrapole moment?Ē Okay? Thatís now down to considerably better than a tenth of millikelvin. Now what is that? Divide that by 3. Thatís three partsÖ

Harwit:

That would be one part in 30,000 or 3 parts in 100,000.

Wilkinson:

Right. So, if youíre willing to interpret the data by averaging over the whole sky, youíre down also to this three parts in 105. But if you want to look at those data and say whatís going on at 6į, then you canít do that because you canít average together all these hours and hours of balloon data.

Harwit:

Sure.

Wilkinson:

So in that range, in this intermediate angular range, our knowledge is much worse.

Harwit:

And you figure thatís where COBE might have a chance to fill inÖ

Wilkinson:

COBE will do better there, yes, but in a somewhat different way. Weíre actually doing some intermediate scale measurements right now. In fact, Peter Timbie is up on the roof, probably, right now doing it. The idea there is to put some beams around the north celestial pole with about 2-3į width and integrate the hell out of the region right around the celestial pole. The reason for picking that is obvious: You can observe it all day long. The problem with large-scale anisotropy measurements like COBE will make is you spend too much of your time scanning over the whole sky and instead of sitting there integrating the devil out of half a dozen spots.

Harwit:

Small areas.

Wilkinson:

Right. The only way to get to microkelvins is to integrate.

Harwit:

So you can do that from the roof here, you figure?

Wilkinson:

Well, thatís what weíre trying to find out right now, so that if the atmosphere gives us trouble, we can beat it out to Colorado this winter, and do it from a dry spot there.

Harwit:

And youíll do that at what wavelength?

Wilkinson:

That is being done at 46 GHz, around 7 mm. That was chosen only because we recently gotten a SIS mixer operating there, competitive with maser noise figures. We had this instrument, and we wanted to do this problem. If you were to start from the beginning and decide what instrument you would use for the problem, it probably would have not been at that wavelength. The atmosphere is not the best.

Harwit:

Now, the work youíve done from balloons has alsoÖas I remember you were the first people to see a dipole anisotropy and, I think, talk about it at meetings. Although, I think, Smoot then published first, and then you published your results, it was sort of a general agreement, though, on the results between the two.

Wilkinson:

Right. Again, the priority, itís hard to establish.

Harwit:

Yes, sure.

Wilkinson:

Certainly the Berkeley data were the most convincing — first, most convincing data. You look at their results and there is no question that the dipole is there. In retrospect you can go back and look at Paul Henryís data, and itís clear he was seeing that; he had the right direction in the sky. He had the right magnitude for it.

Harwit:

Paul Henry?

Wilkinson:

Paul Henry. He was the graduate student here way back sometime; published in Nature. (Note: The reference is given in Appendix 1, item 2.) But only in retrospect would you believe that. At the time it just was not statistically strong.

Harwit:

You didnít publish with him?

Wilkinson:

I didnít publish with him on that.

Harwit:

Why not? I mean he was a student of yours.

Wilkinson:

He was a student of mine, but I thought he did most of the work and deserved to publish it alone. At that point that was the first balloon work. I was so naive that I sent him down to Palestine with this X-band radiometerÖ

Harwit:

(laughter) Oh, the poor boy.

Wilkinson:

Öin a box (laugher) with some Styrofoam around it. I had never been there, and hereís this graduate student showing up there all alone. I think after thatÖwell, obviously he got nothing, although he worked hard and the balloon base really helped him. The next trip I went with him and found out the realities of ballooning. But I felt so badly about having sent him down there the first time that I really didnítÖI wanted him to bask in all the glory of any results.

Harwit:

That sounds reasonable, yes.

Wilkinson:

It was in there. Letís see, Brian Corey and I did get a more plausible measurement of the dipole, and, in fact, were talking about it, at the time Smoot and Gorenstein published.

Harwit:

Yes, I remember that you had come out with a talk before that.

Wilkinson:

They agreed.

Harwit:

Thatís right, thatís right. A lot of times when someone gives a talk somebody else then also is encouraged to publish. It gives you an additional [confidence].

Wilkinson:

Actually, I knew about their result quite a long time before the paper came out.

Harwit:

Is that right?

Wilkinson:

Tony Tyson had spent a sabbatical out there and had worked on this experiment.

Harwit:

Oh, with them?

Wilkinson:

With them.

Harwit:

I see. Heís not one of the people who published with them, I donít think.

Wilkinson:

No, heís not. He published on the instrumentation paper, but he was not in on the discovery paper.

Harwit:

I was just trying to find that paper.

Wilkinson:

That must be in there, Corey and Wilkinson.

Harwit:

You also have Currie.

Wilkinson:

No, thatís Doug Currie.

Harwit:

Yes, okay.

Wilkinson:

What the heckÖthat should be right around time that George published his paper, a little later.

Harwit:

So that would beÖ

Wilkinson:

I donít remember when that was.

Harwit:

Öaround Ď78.

Wilkinson:

Yes, Ď79. Look around Ď79 or Ď80 for Corey. No? Gosh, maybe there is another paper I donít have on my list.

Harwit:

I donít think so. Maybe we should take a look.

Wilkinson:

Itís funny. I know that got published. I made the mistake of letting Brian get away before heíd written his thesis.

Harwit:

What happened then? He didnít finish?

Wilkinson:

It took him a long time to write it, because heís meticulous. Itís a beautiful thesis — the best one. Itís a document; itís a completed thing.

Harwit:

I see.

Wilkinson:

Itís just gorgeous. In fact, George had a preliminary copy of the thesis which was a big help to their work. For heavenís sakes, itís not in there. Well, thereís a paper in Ap. J. Letters.

Harwit:

It is in Ap. J. Letters? Okay, fine.

Wilkinson:

Yes, Corey and Wilkinson, and that gives his number at 19 GHz, I think it was, yes.[5]

Harwit:

Now, at one time a few years ago there was some question about whether there was a quadrupole effect, and I think you then measured the this carefully and decided that there was not, as you just pointed out.

Wilkinson:

Not just us. Phil Lubin also did a careful measurement, and thereís a companion paper with ours where he also points out he doesnít see a quadrupole at that level. Alright, I wish I had a completely convincing straight story for why we got fooled, but I thinkÖ

Harwit:

Well, Melchiorri had published something.

Wilkinson:

Yes, Melchiorri had a quadrupole that, in fact, was not in disagreement. I was not much impressed by that because I knew they were having a lot of trouble analyzing data and that they had had a lot of instrumentation problems in their flight, so I didnít take that very seriously. I think that we got fooled by the Galaxy, our incomplete galactic model. I think there are still some surprises to come from the Galaxy at 1 cm; we still donít understand it well.

Harwit:

Well, sooner or later youíre bound to get some sort of measurement from the Galaxy.

Wilkinson:

And I think there may be a diffuse component of Bremstrahlung that has not been seen by other means, and you canít see it in the radio because itís covered up by the synchrotron emission. Itís already faded out by the time you get to the infrared. The only place to see this would be right around between a centimeter and 3 millimeters, and the data still arenít quite good enough to be convincing there.

Harwit:

Well, itís interesting Wielebinski, in Bonn, was telling me a couple of years ago that he thought that the free-free emission extended quite a bit further into the short-wavelength region in extragalactic sources it was, if Iím not mistaken, than he had expected. Thatís in some work he and a graduate student by the name of Klein had done.

Wilkinson:

Thatís interesting.

Harwit:

You might be able to find that.

Wilkinson:

I found one paper of his where they mapped Andromeda at a centimeter, and that was fascinating. It did look as though there might be kind of a fat component of centimeter radiation in Andromeda even. I remember talking to him, in fact, at this Royal Society meeting. I mentioned this business about model not working very well. At that point I guess weíd already decided that quadrupole wasnít right? I donít remember, maybe, but I was really worried about it. And was worried about a big diffuse component, so I was eager to talk to him. I knew he was an expert on this, and at the time he said that wouldnít surprise him very much and that they were looking at extragalactic sources for it. So, Iíve been looking for his papers.

Harwit:

Yes, I donít know if that is published or not. I almost would have thought it had been, I know he mentioned it to me, I believe a couple of years ago, and was asking me whether I thought from the infrared point of view whether you would expect anything else there, but they seemed to be getting continuum radiation down to shorter wavelengths.

Wilkinson:

And stronger than he would have expected from just HII regions?

Harwit:

Yes, I think that was the point; I donít remember the details.

Wilkinson:

Well, I suspect that that is the problem with that quadrapole. At this point, clearly, we arenít seeing it in the maser data.

Harwit:

Okay, are there any other things that we have missed, particularly on the background radiation? I skirted around a lot other topics that have been of more peripheral interest to you, but wanted to leave this to the end, and Iím not sure whether Iíve covered that, or other things that youíve been interested in which youíd like to talk about.

Wilkinson:

Oh gosh, Iím interested in a whole lot of things not enough time to work on a lot of these things. Well, Iím thinking about another g-2 experiment. Would you believe that?

Harwit:

Really? Yes.

Wilkinson:

I think Iím reverting back to my childhood. (laughter)

Harwit:

Well, thatís always interesting. These fundamental quantum electrodynamic [matters].

Wilkinson:

Yes, and the group at the University of Washington has just done — well, within the last five years which is ďjust,Ē as far as this topic is concerned — this beautiful experiment of trapping one electron and doing resonance measurements on it forever. That triggered an idea — an old idea or dream — to build a Crane-type trap in a cryogenic situation, and actually go in and irradiate that thing, flip the spin with radiation so youíd directly measure the spin flip frequency and change the Landau level directly with radiation. You would end up measuring g with a ratio of two frequencies, which is what you really want to do. The Crane technique still involves having to measure and average the magnetic field, but there really isnít any reason to have to do that, if you do it right. It also has always intrigued me that very little work has been done where you put electrons in their lowest possible energy state, get those Landau levels populated the way you want them. Also, I think, this thing could make an absolutely marvelous infrared detector.

Harwit:

How would it work, with a singly electron?

Wilkinson:

No, in that case, just to get the absorption cross section large enough youíd want to have a lot of electrons, and then you would essentially study the distribution of electrons in the states. Youíd pump them into a nonthermal state, expose the thing to the radiation — I would be interested in thermal radiation at half millimeter or millimeter wavelengths — and watch the distribution change back to thermal.

Harwit:

Very nice.

Wilkinson:

And the whole thing is down there at, you know, three-tenths of a Kelvin, say. Itís kind of the ultimate thermal detector, it seems to me. Thatís a day dream.

Harwit:

Another decade?

Wilkinson:

Right. (laughter) Probably. Another thing Iíve been thinking about for several years which is a nice astronomical topic, I think, and have had seniors working on — itís at that level — is to go look for low mass stars, below a tenth of a solar mass by looking at red dwarfs and doing classical spectroscopic binary observations on red dwarfs. If you look at the statistics of how low mass stars — dwarfs — behave, a lot of them are in binary systems, and so why not, you know things between Jupiter and the red dwarfs.

Harwit:

Okay, so the fact that itís a low mass star would give you a somewhat higherÖ

Wilkinson:

Ösensitivity, yes, mass sensitivity, right.

Harwit:

Ösensitivity to a Jupiter-sized object.

Wilkinson:

A companion. But itís always puzzled me that there is this gap between Jupiter and a red dwarf that we know nothing about. It would be fun just toÖ

Harwit:

We donít have one locally.

Wilkinson:

Well, we might, who knows? Not real locally.

Harwit:

Thatís sort of almost the Barnard star problem.

Wilkinson:

Yes, exactly; it is.

Harwit:

Which is a tough one.

Wilkinson:

Not easy. Actually had a senior a few years ago who built an instrument and put it on our 36-inch and measured some spectral lines; he used Hα which is an emission line in dwarfs.

Harwit:

You have to do this for decades, wouldnít you?

Wilkinson:

Well no, no, thatís another nice thing about dwarf stars is that youíd look for nearby orbits — monthly, maybe yearly orbits.

Harwit:

I see. Youíre saying you think that the Jupiter-sized object would be an astronomical unit away?

Wilkinson:

Why not?

Harwit:

I donít know if it would keep its hydrogen. I suppose it would. The dwarf wouldnít put out that much radiationÖ

Wilkinson:

Yes, thatís right, itís in a relativelyÖ

Harwit:

Öand the gravitational attraction might be large enough. I donít know if youíd form them thereÖ

Wilkinson:

That might be a problem.

Harwit:

The core of Jupiter isnítÖitís more massive than the Earth but not that much more.

Wilkinson:

Formation is a real problem. Iíve not looked into that as much as I should, and it might be a completely stupid idea. There may be reasons why you wouldnít form close dwarf binaries.

Harwit:

But if you had good spectral resolution, it might be that you could get some of these things faster than the old-timers did.

Wilkinson:

Oh, you can do much better, much much better. This student, incidentally, is somebody youíre going to hear from, if I am any judge of people. This guy was amazing. His name is Andy Lange. Heís now working with Paul Richards on doing a spectrum of the background radiation. But heís really super. He invented a spectrometer, built the instrument, had a CCD in it, took it to the telescope, observed, wrote a beautiful senior thesis, didnít find anything — but not too surprising; he had only a couple of months to observe, all for his senior thesis topic.

Harwit:

Thatís pretty good. What sort of an instrument did he invent?

Wilkinson:

He ended up taking an interference filter and using it like a Fabry-Perot, but a Fabry-Perot in a converging beau, and that was a new idea to me, so that spectrum ends up being a set of rings — aligned. It gives you a bright rim in a sort of donut-shaped continuum. Itís not terribly efficient, but given a CCD detector, which is what he had, this and an interference filter lying around; it worked fine. A very ingenious guy.

Harwit:

One thing I forgot to ask you, in closing still, you know I told you that one is interested in daily lives of scientists: Do you have hobbles? I met your daughter at one time.

Wilkinson:

Thatís right.

Harwit:

Thatís the only child, is that right?

Wilkinson:

No, I have a son also.

Harwit:

Do you? I see.

Wilkinson:

Yes, In fact, you met my daughter twice: At your house, and then getting on an airplane, heading for Greece.

Harwit:

Thatís right at Kennedy.

Wilkinson:

Thatís right. I have a son. Heís now enrolled at Colorado, senior year.

Harwit:

Is he interested in science?

Wilkinson:

No.

Harwit:

Your daughter is interested in French, I think, wasnít it?

Wilkinson:

Sheís a sculptor.

Harwit:

I see. I donít know why I thought she was in French. Where is she now?

Wilkinson:

She may have been speaking Italian at the time. She learned Italian.

Harwit:

She was at Wells?

Wilkinson:

Thatís right. She was at Wells, transferred to Syracuse after three years, finished up at Syracuse because they had a very good sculpture program. She now works for the Johnson Atelier which is here in town, well, half-way between here and Trenton; makes these bronze realistic figures, reading newspapers and skateboarding and so forth.

Harwit:

Oh really. Oh yes, I like those, yes.

Wilkinson:

Yes, theyíre quite nice.

Harwit:

They take you aback, yes. There is the one you have here at Princeton?

Wilkinson:

There are several around town, the guy sitting reading a newspaper down in Palmer Square.

Harwit:

Yes, thatís right.

Wilkinson:

That was one of the first.

Harwit:

Yes, then he has some nowÖthere is this one in New York City of the man hailing a cab.

Wilkinson:

Oh, I hadnít seen that one. There a number of them around. Thereíre many in Washington.

Harwit:

You always have to take a double take.

Wilkinson:

Theyíre so realistic.

Harwit:

Yes.

Wilkinson:

Anyway, thatís kind of their bread and butter, and they then also have a big foundry, a whole operation for making bronzes. They have well-known artists who come in and have their pieces done there. Wendy was an apprentice; sheís now on their staff, and does work of her own on her own time. Sheís quite happy about that.

Harwit:

And what will your son do?

Wilkinson:

Heís interested in writing and history — not clear. At the moment heísÖ

Harwit:

Itís a rough career.

Wilkinson:

Öa baker. He makes his living by baking.

Harwit:

While he goes through school?

Wilkinson:

Yes.

Harwit:

Youíre married, also?

Wilkinson:

At the moment not, but soon to be remarried.

Harwit:

I see.

Wilkinson:

I am divorced from Wendy and Kentís mother. She lives here in town as well.

Harwit:

Their mother?

Wilkinson:

Yes.

Harwit:

Good. Have we forgotten anything?

Wilkinson:

Oh, well, lotís of things. I donít know what kinds of things youíre interested in. What kind of support or motivation or whatever, and you started talking about family and influences and so forth, you need. I really think itís incredibly important to have a supportive and understanding family to be able to do these kinds of things.

Harwit:

Have you felt youíve had that?

Wilkinson:

Oh yes.

Harwit:

This is because the family gave you the time to do your work.

Wilkinson:

Yes, they were very understanding about working nights and spending a lot of concentration on your work. Maybe not having vacations as often. I can remember several summers when we first came to Princeton, we didnít take a vacation. Itís really kind of unfair for a family to have to put up with that, but I think itís people who do the kinds of things we do, who donít recognize that it does affect their families, particularly, and more so than in a lot of professions.

Harwit:

Yes, I was going to ask you whether you felt that it strained your family life.

Wilkinson:

I think it did.

Harwit:

It is tough, I think.

Wilkinson:

I could have, in looking back nowÖ

Harwit:

Was your wife interested in science?

Wilkinson:

No, no. In fact, she was a hometown girl. We met while we were still in high school.

Harwit:

I see.

Wilkinson:

Got married while I was still in graduate school.

Harwit:

What does she do now?

Wilkinson:

Sheís getting her masterís degree at Rutgers inÖI always get this wrongÖ

Harwit:

Approximately.

Wilkinson:

Well, sheís a social worker and deals with older people. Now whatís the work for that?

Harwit:

Geriatrics?

Wilkinson:

Geriatrics, okay. I always go through all of the other things that begin with ďgĒ. Sheís interested in care for the elderly.

Harwit:

Thatís an interesting thing now. There are more and more people getting old.

Wilkinson:

Very important field, and itís obvious that weíre pretty much mistreating older people — not taking care of them more, paying attention. Sheís made quite a good life for herself.

Harwit:

Did she go into that after the children were grown up?

Wilkinson:

Yes, and after the divorce.

Harwit:

Oh, after the divorce? So is that a recent divorce?

Wilkinson:

No, we separated 10 years ago, 12 years ago.

Harwit:

Oh, I see. So, she really had a chance to go back to school then?

Wilkinson:

Yes, yes, started a career.

Harwit:

And the children, then, were perhaps 10-15 years old?

Wilkinson:

Thatís right. They were both still in school, and we worked out a very good custody arrangement six months on, six months off. It worked out well.

Harwit:

Do you think that your working as hard as you did, decided that there would be a divorce in some way?

Wilkinson:

That contributed to it.

Harwit:

Itís certainly is very difficult. I know from my own experience that itís no fun on wives when you go off to a rocket range for a few weeks, and the furnace breaks down in the middle of winter, and all those things. Itís very hard.

Wilkinson:

Yes, and the kids look around and see that other kidsí dads are going to the beach and their dad isnít. Itís tough. Itís as hard on the kids as anybody else.

Harwit:

But your relationship certainly with Wendy — I donít know about with your son — seemed to be outstanding. I remember running into you when you were going off to Greece together.

Wilkinson:

Oh yes, oh we had a great time. I enjoy the kids very much, and go out to Colorado to ski with Kent. Heís a very good skier.

Harwit:

Well, thatís nice when one does things together like that, at least occasionally. Itís better than nothing at all, and I think the children prize it to some extent.

Wilkinson:

But one does have to ignore a lot of the other parts of enjoying things and living, in order to do this kind of thing, unless youíre so brilliant you can think about it three hours a day and go home or something.

Harwit:

Thereíre very few people who do that, though.

Wilkinson:

It seems that way, though, doesnít it?

Harwit:

Itís a fairly spartan existence, I guess.

Wilkinson:

Well, I must say, the last 10, 12 years since Iíve been pretty much either living with a couple of the kids or alone, have been fine, and I can work as much as I want to. I enjoy that.

Harwit:

Do you find that you work yourself to a point where you donít really think straight anymore, and youíve got to force yourself to relax, but thereís nobody there to force you to do it.

Wilkinson:

Yes, right. Thatís the biggest drawback.

Harwit:

I get that on trips. You know, if Iím away for a month someplace, I mean, itís not as though you could kind of work all the timeÖ

Wilkinson:

No, thatís right, you canít.

Harwit:

Öand you donít have any regular relaxation patterns that other people might have, so you just keep on working until you almost exhaust yourself.

Wilkinson:

Until you drop. Thatís very inefficient. You donít get very many good ideas that way.

Harwit:

How do you get around that?

Wilkinson:

I usually save mundane lab work for late in the afternoon or when Iím tired; try to do anything I have to do where I need to think in the morning. That helps.

Harwit:

But you donít read or listen to music, go to movies, or anything like that?

Wilkinson:

I enjoy music a lot. Every once in a while I get to a point where I just have to go and hear some live music. Thatís a big help. Get on my motorcycle and take off across the countryside. This part of New Jersey is beautiful, and there are still back roads.

Harwit:

I noticed it driving this way, yes.

Wilkinson:

It can be very, very pretty. There are a lot of little back roads that go off, and you know, on a motorcycle all alone you really can relax.

Harwit:

But you donít do that in winter, I imagine.

Wilkinson:

Well, no, itís kind of rough in the winter.

Harwit:

Yes, thatís what I thought.

Wilkinson:

I go skiing in the winter.

Harwit:

Yes, okay

Wilkinson:

Go up to your place — Greek Peak.

Harwit:

Have you been up there a few times?

Wilkinson:

Yes, yes, when Wendy was up there.

Harwit:

Yes, sure, itís close.

Wilkinson:

I used to come up and go skiing with her. I enjoyed that place. We always had a good time there.

Harwit:

So when are you getting remarried?

Wilkinson:

Can you keep a secret?

Harwit:

Iím sworn to it by this machine. (laughter) If I turned it offÖ

Wilkinson:

Is this machine still on? Oh my goodness. Turn it off and Iíll tell you.

Harwit:

I can turn it off.

Wilkinson:

October 13.

Harwit:

Very good. Oh, that close. Congratulations.

Wilkinson:

But please donít say a word. Even my friend next door doesnít know about this.

Harwit:

I will not say anything, and this is probably

Wilkinson:

Itís a second marriage for both of us so weíre trying to keep it very low-key.

Harwit:

This is probably as good a time to turn this off asÖ

Wilkinson:

Fine.

Harwit:

Thank you very muchÖ

Wilkinson:

Youíre welcome.

Harwit:

Öand if anything occurs to you overnight, and I think you wanted to look something up — we discussed up earlier on that I canít rememberÖ

Wilkinson:

I was going to look for a notebook to see if there were any dates that might help you decide when something happened.

Harwit:

Yes, that would be nice. If you still can find something before I leave sometime tomorrow, we could just record that still.

Wilkinson:

Okay, fine. Iíll go see if I can find some notebooks.

Harwit:

Thank you very much.

Wilkinson:

Good.

[1]Wilkinsonís first notebook on the background radiation problem is dated December Ď64, but apparently the notebook was not started right away, because he was already mapping his horn on December 29, 1964. Peter Roll, at the University of Minnesotaís Physics Department, might have earlier data in his files, Wilkinson thinks. The project, he believes, must have been in action a couple of months by the time because of the preparations required and mentioned in the interview. In January 1965 Wilkinson was continuing the mapping of the horn. He thinks that he must have started in on the project in October of 1964.

[2]Missing page.

[3]M. H. coincidentally had an opportunity to talk with Dr. Joseph Weber about this on the evening of September 30, 1984 — just three days after this interview at a reception in celebration of Edwin E. Salpeterís upcoming 60th birthday. Weber recalled he had approached Gamow in 1948 about doing a Ph.D. thesis on a topic that would make use of Weberís wartime background in microwave engineering. Gamow reportedly replied that he had no such topic.

[4]Missing page.

[5]Coreyís results were published with the paper by Cheng, Saulson, Wilkinson and Corey, Ap. J. Lett., 232, L139-143 (1979).