Oral History Transcript — Dr. Bengt Stromgren
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.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Bengt Stromgren
Bengt Stromgren; May 13, 1976
ABSTRACT: Family background and early interest in astronomy (Elis Stromgren). Undergraduate and graduate studies at University of Copenhagen, late 1920s; studies at Niels Bohr Institute, 1927–1929; thesis work in classical astronomy (orbits of comets). Development of photoelectric photometry and observations; early electronics, 1925; conflicting results in calculations of opacities (Arthur Stanley Eddington, Gaunt, Thomas T. Sugihara, Svein Rosseland, J. R. Oppenheimer, Meghnad N. Saha, R. H. Fowler, E. Arthur Milne, Cecilia Payne-Gaposhkin); assistant at University of Copenhaagen, 1929; LaSilla Observatory. To University of Chicago and Yerkes Observatory (Otto Struve), 1936–1939; starts work on formation of H #II regions, 1939; work at Mt. Wilson Observatory on absorption lines (Walter S. Adams, Theodore Dunham); estimates of ages of stars (Hans Bethe); Hubble Constant; comparison of astronomy in Europe and U.S.; European astronomers in U.S. (Gerard Kuiper, Polydore Swings, Carl Osbourne, Ejnar Herzsprung); comments on history of Yerkes (Struve), teaching at Chicago; discussion of work on equation of ionization and calculations of opacities (Carl von Weizsacker, Struve, S. Chandrasekhar); comments on W. W. Morgan. Discussion of work on stellar evolution, ionization of interstellar hydrogen (Struve). Effects of World War II on astronomy; influence of European astronomers on Americans; Ludwig Biermann; European Southern Observatory; developments in astrophysics during the war; optical studies. Nazi occupation of Niels Bohr Institute (Werner Heisenberg); contacts with German astronomers; development of Brorfelde Observatory. Becomes director of Yerkes and McDonald Observatory (Robert Hutchins), 1950-1957; American astronomy after World War II; relation of scientific community and government (Office of Naval Research); stellar classification work. Institute for Advanced Study, Princeton (Oppenheimer), investigation of intermediate population II and extreme population II, 1957; establishment of Kitt Peak Observatory; return to Denmark, 1967. Also prominently mentioned are: Niels Bohr, Werner Bolton, George Ellery Hale, Jacobus C. Kapteyn, Lev Landau, and Harlow Shapley.
Session I | Session II
Hoddeson:This is Lillian Hoddeson and Gordon Baym and we are carrying out the second session of an interview with Professor Bengt Stromgren in Professor Stromgren's office in the Copenhagen Observatory. Today is May 13th, 1976; it's shortly after 10 a.m. in the morning. We left off last time, in the middle of your stay at Chicago.
Stromgren:Yes. We did talk about the first six months, when I was teaching on the campus. One of my activities, in addition to giving courses, was to put the finishing touches on some HANDBUCH articles that I had written, and I got involved with (pointing to reprint) — I thought you might like to glance at that — that's what I was doing in those days. That was for the handbook. The College Handbook.
Hoddeson:I've seen this article in the library.
Stromgren:And I was also the editor of the whole volume, but I wrote those two chapters, and then shortly before that, I had finished these HANDBUCH articles. So these activities had given me an opportunity to read up on the whole literature, up to that time, the middle thirties, in the fields in question, photometry and this (pointing) was stellar interiors, stellar atmospheres.
Hoddeson:The HANDBUCH article appeared in 1937. This one is 1936.
Stromgren:Yes. That had been finished before I left Copenhagen, and this, I was reading proof. The first six months in Chicago, that we're talking about.
Hoddeson:In the Preface you state that Hans Rosenberg originally planned the volume.
Stromgren:Yes, that's right, when he was still working in Germany. And then he left, because he left in the early period of the Nazi regime, when it was still possible to take things with you, and he then accepted an offer from the University of Chicago. But since he left Europe, he couldn't continue his activities editing this, and I took over.
Hoddeson:Your father also played a role in this volume.
Stromgren:Yes. He knew Rosenberg. But he didn't — I think that's explained in the Preface — he didn't do the editing himself.
Hoddeson:Did you also edit the other articles?
Stromgren:Yes, in that whole volume. It was, as I said, an opportunity to read a lot, for a half a year or so.
Hoddeson:How did this affect the way you went about your work at Chicago?
Stromgren:Actually, this was a chapter that had been finished. I was just doing the last bit of work on that. One problem that I was working on, during those six months in Chicago, had to do with stellar atmospheres, and it was connected with the efforts to understand the damping, particularly in solar type stars. Whereas if you look at this — that was written in 1935 — you see that it has already begun to be clear that the damping which was determined empirically from curves of growth, for solar atmosphere, was due to neutral hydrogen, but the whole problem wasn't understood, until the role of H- was discovered by Wildt a couple of years later, because that thing explained why the hydrogen pressure was a large as it had to be to explain the damping.
Baym:Had you considered this yourself, as a possibility?
Yes, this is in here, the suggestion that it might be damping by hydrogen. But there were more and more certainly factors of 10 or so missing in the story, and that's because the absorption was wrong. It was computed as absorption by metals, and it was only when it was computed as absorption by metals, and it was only when it was computed correctly, absorption by H-, that everything came out right. From then on, it's been [ ] improvements of say, factors of 1 1/2 or 2, but not these gross discrepancies that we had. So, that is part of my work, during those six months. But as I say, it was a very exciting time in the life of the university, and we were exposed to many new impressions during those six months. Nevertheless, I welcomed the change, to go to Williams Bay and there, we got there in April, 1937, and there I worked more in the direction of stellar interiors. I had to write, and I did this in July, an article for the ERGEBNISSE DER EXACTEN NATURWISSENSCHAFTEN. I agreed while I was still in Copenhagen, with Hund that I would write such an article, and this I did. And in that connection I did some work in continuation of earlier work on the interior composition of the sun and other stars, for which well-determined masses, luminosities and radii were available. My first work on this problem that we have discussed was based on new calculations of opacities and mean molecular weights, that had been based on calculations according to the equation of ionization.
And the result I came up with was, roughly, one-third by weight, hydrogen, and two-thirds metal. But, now, what I investigated was, what change would be made if helium was added? And the final result was a composition that is remarkably similar to the one we've got today, which means I found four percent by weight metals, and between 60 and 70 hydrogen, and the rest helium. Now, it came out that — well, it was just luck, because helium now has a free parameter, for the sun. It was just clear that it might be an abundant element. The only observational basis then was on the relative amounts of hydrogen and helium in prominences — but the determinations were pretty weak, and the prominences might not be typical. The reason I took it seriously was that, particularly, von Weizsacker had discussed what he called the building-up hypothesis, the alpha-particle hypothesis, and concluded that the abundance would give a certain ratio of helium to heavy elements, which I assumed. Now, the reason was, all along, in those early days, when there was no nuclear physics to speak of to support the reasoning, that the ratio was correct.
Baym:Did you assume that the abundances were the same in the atmosphere and in the interior?
Stromgren:No. There wasn't, at the time, sufficient information on the atmospheres to use that at all. You remember, this was just before the H- breakthrough, and our knowledge of the atmosphere composition came only a little later. Now, what was done then was to — and this, you can follow from the forties and early fifties — was to assume that the abundances given in the atmosphere — now there are fairly reliable compositions, and by the early fifties, particularly, it was also clear that there had been complete mixing of the whole sun, in an early period, and that in any case, there was a de-particle zone. But as I say, this came later. At the time it was just an assumption, and luckily came out right. But the main thing, I think, was that much pointed in the direction of a substantial helium content, and that that would then reduce the metal content to a few percent. And that of course had proved to be right.
Hoddeson:Who did you discuss your work with at Chicago?
Stromgren:Well, in that year particularly with Chandrasekhar and with Struve, who were both working at Yerkes Observatory in Williams Bay. And particularly my discussions with Chandrasekhar were very important for me. He started writing his big monograph on stellar structure and we discussed chapters as they came along, and had almost daily contact during that whole stage, the first time around, at Yerkes until April 1938, when I left. Now, another part of my activity was participation in observations with the 40-inch refractor, and this was done in collaboration with W.W. Morgan, who, at the time, was developing his system of classification that has become a prominent part of the whole story of stellar classification. He had it almost ready in his head by that time, and was taking a lot of spectra. In the history of astronomy, there are off the record stories about disputes between Mount Wilson and Yerkes at the time, but you can find that out, perhaps in some other way. They didn't like it at Mount Wilson. Adams wrote a strong letter to Struve saying that this young man Morgan better leave off. Well, this story was referred to last summer, there was an IAU symposium in honor of Morgan, and a number of the old stories were revived at that time.
Hoddeson:Did you interact with Morgan personally?
Stromgren:Yes. In fact, in addition to these discussions with Chandrasekhar and Struve, that were so valuable, the discussions with Morgan were very valuable to me, and influenced my attitude to stellar classification and such problems.
Hoddeson:Do you remember any specific incidents, that you think would give a picture —
Stromgren:I would say, in a general way, this was the first time I had occasion to look at many good spectra. They were in the files, and Morgan discussed problems with me. And we worked on a fairly modest program together. It was of some interest. I took, as part of that program, plates to supplement his own. They were all plates of cepheids. We wanted to examine the relation between the period of the object and the spectral type, according to this new classification. We came out with very good results. But for me, this was a very interesting experience. It was the first time I observed with a big instrument. Not a very convenient one, but one that had been adopted by Morgan and streamlined so that that particular type of classification spectra was the type of material that was very good indeed.
Hoddeson:When you say "in the files," do you mean — ?
Stromgren:All the big observatories, particularly the American observatories, had and have files of plates. Each observatory has its own. Now, there are centers, because such organizations as ESO and before ESO, the Observatoire de Haute of Provence had to worry about that; the Observatoire de Haute of Provence instituted a system that they call the [ ]. That's a library of plates. And this had to be carefully organized, because a number of people go there from other places, and the idea is that they use their material, and when they have finished, it goes to the basic library, and can be referred to by other people who might later on be interested in that particular star.
Hoddeson:How far do each of these files go?
Stromgren:The most famous plate collections are those of the Harvard Observatory, direct plates, sky surveys, and even today they are useful. They start in the 19th century, and so, for instance, when the first quasars were discovered, and identified as such, variability back to those days could be established. They were also used for compact x-ray sources. They will continue to be a useful source of information.
Baym:You were also starting in this period to think about stellar evolution.
Yes. Actually, even before I left for America, my first paper on the hydrogen content of the sun and stars had been followed by an attempt to see if there were systematic trends in the chemical compositions. There was some evidence for this, and what one expected was a gradual change from hydrogen to — well, some people thought metals — or to helium — so, a diminution of the hydrogen content. Now, it wasn't clear then whether or not one should assume mixture of the whole star, and the tendency was, on the basis of some upper limits for the mixing currents, derived by Eddington, to assume that there was mixing. Remember also that, as we discussed in the first interview, the first time scale that was thought of was a good deal shorter that what we think of today. Now, in view of this it was perhaps natural just to look for systematic changes, diminution, in the hydrogen content, and to try to correlate that with location in the Hertzsprung-Russell diagram. I tried to use the material that was available, and found such effects. But when we look at that today, you see that perhaps there were some indications that go in the right direction, but we knew far too little about what the actual changes are, and particularly the whole approach to the effects of changes of chemical composition. We only formulated it after it gradually became clear that there is mixing, at least for the sun and along the main sequence for the core, and again, that mixing is not important for large parts of the outer star. All of this only became clear gradually through the forties and early fifties. But by the early fifties, it was quite clear what the situation was. As you can see from, say, this article on "The Rise of Astrophysics," which goes up to that time — by then, we knew what the composition was.
We also knew what the models in the main sequence were. In particular, that you have a rather deep outer convective shell, to around F0 and that when you got to the larger masses, the A stars, you do not have this outer convective shell, but you have instead a convective core. By that time, the stellar interior models on the main sequence were well established, and this was of course the starting point for seeing what the effects of the changes in chemical composition would be. I think, when you look at the history, it's important to realize that two pieces of information were necessary — what happens in each volume element to the chemical composition, and what the mixing is. Now, I referred to Eddington's upper limit. And what happened was, it was realized in the forties and fifties that there was an upper limit, all right, but it was far above the actual mixing rate. When that was calculated, I think the final answer was given in a publication by Sweet in England. Then there was no doubt any longer that, on the main sequence, except in exceptional cases of high rotation, mixing doesn't play a great role. And you just have that the convective core is mixed, and this then is the recipe we need. Now, it still took a while till, in the fifties, the evolution from the main sequence into the giants was understood.
Baym:One question here — you had thought about the presence of helium in stars, and also stellar evolution, but had it occurred to you that hydrogen would evolve into helium?
Stromgren:Well, I think once Bethe's work had been published, this was clear. But when I first worked on that, it was pre-Bethe, and at the time when one still thought of the possibility of immediate further build-up.
Hoddeson:Was that the general position?
Stromgren:No. There was no general feeling, no. It was all groping in the dark, until the time of Bethe's paper.
Hoddeson:Were there seminars on it?
Stromgren:Yes, there were. For instance, as I mentioned last time, the meeting in Washington that was organized by Gamow and Teller, and where Bethe was present. Things still were not clear. But it didn't last so long after that. And before that, I had quite a bit of correspondence with von Weizsacker, who gradually came quite close, as you know.
Hoddeson:Does that correspondence exist?
Stromgren:I'm afraid not. I haven't been very careful, and because of my travel, I have lost ...
Baym:Another portion of your work was on the binary, epsilon aurigae.
Stromgren:This was a by-product of my stay at Yerkes. The basis was some work that Kuiper had done on discussion of the radial velocity, and the data on the eclipses. This led to a very large parameter for one component, and it was quite hard to explain the light curve, because it looked as if, although this light curve had the shape of a total eclipse, the sunlight would come through. Now, under normal circumstances, you don't have a constant light in the minimum, and at the same time, the star shining through. Berkustra and I contributed the spectroscopy, tried to understand this. What I did was to compute what happens when the small star shines through the outer atmosphere of this very big tenuous star. And the simplest form of the calculation, when gives the simple answer, is one that shows what happens if the small star ionizes the atmosphere of the big star and frees electrons. These electrons then scatter the light of the small star, so that the opacity that reduces the light of the small component is produced by that small component itself. Under those circumstances you can have a constant absorption, although the path through the atmosphere of the big star is strongly variable. Well, this was just calculation, my contribution was just calculation of that sort of model. As far as it went, it was all right.
Hoddeson:What was the nature of that collaboration? How did you work together?
Stromgren:Well, Kuiper discussed the orbit, and the eclipse observations, and on that basis, he produced the model of a big star and a small star and certain radii for the two, and Struve discussed the evidence from the variation of the spectrum. What I did was just to contribute this schematic model, what happens if the small star shines through the atmosphere of the big and produces the opacity?
Hoddeson:Who wrote the paper?
Stromgren:Well, we wrote it together, each his section.
Baym:Struve makes a remark in his book on 20th century astronomy, that outside of [ ] no absorption lines are seen. In 1937 three Yerkes astronomers, probably independent of one another, arrived at the same conclusion. The large eclipsing body, referred to as the I star, consists of a great tenuous mass of gas, whose bulk is partly transparent to optical light, etc. What I was wondering about is his remark: "probably independently arrived at the same conclusion."
Stromgren:That is a cryptic remark. I didn't worry too much about the model. I just made these calculations. Now, let me follow this up. The conclusion was that you had a very large star, and that it must radiate in the infrared. Now, already at the time attempts were made to find this infrared radiation. Nothing was found. And gradually, as the infrared technique improved and improved, this was repeated, and it was never found. So the result, from 1937, really couldn't be right ... And that's the situation today. There must be other explanations. There have been many contributions in the fifties and sixties to this. The latest thinking has of course centered on the possibility that you have a black hole in the system. If you see gravitational effects without seeing a trace of the star, well, that's the direction of your thinking. And I'm not too well-informed on the status of this today.
Hoddeson:Were unusual objects, such as black holes and neutron stars an issue in the thirties.
Stromgren:Not from the observational point of view, yet. There was nothing to go by, except, of course, you go to Baade and Zwicky's paper, and then later than this, Oppenheimer and Volkoff.
Baym:Were you aware of the Baade-Zwicky paper at that time?
Stromgren:Oh yes, and I also knew Landau's thoughts. He was in, Copenhagen at the time. And as we were briefly mentioning last time, in connection with the debate between Eddington and Milne —
Hoddeson:— we have to go back to that.
Stromgren:— quite a few people had an open mind on this, thinking that they could start the integration from the surface in a different way than was done on the Eddington model. It would look as if you could never manage to satisfy the central conditions, but something might turn up, some big deviation from the accepted equation of state, so in the end you would be able to remove the excess mass. But it was — the probable fate of such models —
Baym:Was Landau's paper taken seriously?
Stromgren:Well, it was discussed at the Bohr Institute. I don't think that many of those who worried about stellar structure really took it seriously. But it was kept in mind as an idea, and also this idea that with very high densitities, you could have a big contribution again from contractional energy. I think it was a little more difficult to accept. Nuclear energy was new and fashionable, and superficially there might be — also emotionally — there might be a little bit of the reaction, "Oh no, not contractional energy!" Much of the progress I think can be hindered by that sort of inhibitions.
Hoddeson:Let's see, the Milne-Eddington debate, I don't think we've dated that.
Stromgren:Yes. It is around 1930. And it's all in the MONTHLY NOTICES and THE OBSERVATORY.
Hoddeson:Did you take an active part?
I did one paper. Yes. You see, it was a very simple problem. Milne had made this suggestion that if you start with models, then maybe you could still satisfy the central conditions. And he had gone further and said, "Well, you know that you'll have the conditions in the white dwarf, that is the polytrope 3/2 will be important to the 5/3. And maybe this will remove the difficulty that there is an excess mass." Well, that didn't sound right, that you could remove mass with the law of P equals four or five thirds. But Milne had left it at that, and what I did was to investigate the boundary conditions, and fit, in other words, two polytropes together — N equals three in the outer part, and three halves in the interior. The answer was quite clear — this wasn't an answer, it didn't lead to any model of the kind that Milne had in mind. And I wrote that up, for once very fast, sent the manuscript three weeks after Milne's paper. But I sent it to him, and he was very kind, although it didn't confirm his hopes. He presented it to the RAS [Royal Astronomical Society] as you can see in the OBSERVATORY. And that was that. It appeared almost simultaneously with two other papers that proved the same thing in different fashion, and they were by Henry Norris Russell and — well, I'll come back — and Cowling. So, this was my first effort.
Then, I had a paper where you give up the idea that the outer part has to be a polytrope N=3. It was now clear that the energy production must be strongly concentrated to the center, so what I did was to compute numerically point source models, and I did that with some hopes that the miracle would happen, that as you integrated, you might find that you could fit something — because there were known effects. First of all, we knew that at sufficiently high density there's a transition to the relativistic equation of state, and also that radiation pressure vanishes, because the opacity goes down. And particularly the vanishing of the radiation pressure might perhaps lead to this removal of — The result was negative. In this paper I developed the method of [ ] going through the outer part of the star, where L(r) is constant, and where M(r) is still very nearly constant, but you get through a good part of the star and start your integrations. In those days, long before the electronic computer, such techniques were important — and they were used for some years by other people too — and integration was done for the interior —
Baym:— how did you do the interior, by desk calculator?
Stromgren:Yes. With a small desk calculator. I never found that difficult, and I even liked that sort of work. You work for days and weeks just step by step integration. But I was quite familiar with the numerical techniques, as explained, from my previous work in classical astronomy.
Baym:Did you ever become involved in programming computers?
Stromgren:No. I've always been spoiled, in the sense that the three big computer programs that I used when I was at the Institute for Advanced Study in the fifties and particularly the sixties were developed at the Institute for Space Studies, where I worked with a small group of programs. And this goes both for the stellar interior program, where Kelso and I worked together on rather extensive calculations of [ ] evolution. And it also goes for the stellar atmospheres programs, for B stars. The programming was done by the programmers, who worked with me, so that when I wrote down the equations, I didn't have to do the programming myself. And the third program was the same story. That was on galactic orbits.
Hoddeson:Struve attributes the origin of your work on ionization of interstellar hydrogen to a conversation with you in 1936, while driving to Chicago from Williams Bay. Do you have any recollections of that conversation?
Yes. I remember that we sometimes drove together to the campus, that we always had interesting conversations. My whole interest in the problems of interstellar matter was revived during that stay at Yerkes, and this was largely through the conversations and discussions with Struve. He had — and these were the subject of our first discussions — worked on the particle medium, on reflection nebulae, but when, as we were talking about last time, he made the very important discovery of the Balmer line glow in extensive regions of the galaxy, this triggered my new interest in the questions of the interstellar medium.
Now, there is a connection between the model I worked out for epsilon aurigae and this work, because there's a somewhat similar calculation that you do when you have a high temperature luminous star embedded in a uniform hydrogen medium, that the star produces the ionization that is connected with the opacity. In this case it's the opposite. It prevents the Lyman absorption, as in the other case it produced the electron scattering. Electron scattering in interstellar space doesn't play much of a role. So the game was to follow what happened to the Lyman absorption.
But there was a certain connection, because in both cases you had a source that influences the nearby or surrounding medium. Now, I continued the work on interstellar matter the first months after my return to Copenhagen, the first year almost, and I continued my discussions with Struve on the question. The result was a paper on the properties of interstellar matter that discusses not only the hydrogen situation but also both absorption and emission, also some aspects of the interpretation of the interstellar absorption lines.
Hoddeson:How did Struve function generally as head of the department?
Stromgren:Well, he had enormous drive, you see; he did many things at the same time. He was an extremely productive scientist, and at the same time, he built up the department in collaboration with Hutchins, and he also built up the ASTROPHYSICAL JOURNAL in that period. This was also the period when McDonald Observatory was developed. So there were huge scientific and administrative questions there. Struve wasn't a man who delegated readily, and he did all this himself, so in those days, it's very impressive that he could do all that.
Baym:Did you feel that his administrative responsibilities cut into his scientific productivity?
Stromgren:Never. He managed to produce more than most people. And I don't see that he could have done much more, had he concentrated on that work alone. And indeed, his administrative work, in connection with the University of Texas problems and McDonald, paid off. His work gradually was centered on the use of the cassegrain spectrographs of the 84-inch, where he did a lot of the work that has been the most valuable on spectroscopic binaries and other variable stars, groups that he could investigate with the more powerful spectrographs, and guided by very fruitful ideas that he had.
Hoddeson:Was there much social as well as scientific contact between you and Otto Struve?
Yes, we became close friends and this developed further when, after the war, he visited Europe and Copenhagen, and when I came back. What Struve had intended was to persuade me to go back to Yerkes. I left in April, 1938, and in the fall of 1938, he had worked out a proposal that had the approval of the dean and the president. He then thought that we would share the responsibilities — I would be director of Yerkes and he would be director of McDonald and chairman of the department.
I was very much tempted. It was a time, as you remember, when the future in Europe was quite dark. It was after Munich, and one could see in which direction things were going. So it was a very difficult decision for Sigrid and myself. But in the end, we felt that if there was a catastrophe in Europe, we would rather be there than away. And so I didn't accept the offer then.
Baym:You had an offer from the university here of a full professorship?
Stromgren:Well, when I returned (to Copenhagen) in 1938 in April, after a year and a half, it was to take up a new task — what was called in those days an extraordinary professorship had been set up for me, a personal professorship. So I was professor without any duties as director of the observatory, until 1940, when my father retired, and I then switched to the so-called ordinary professorship.
Baym:While you were in the States, had you not been offered a professor position in Copenhagen?
Stromgren:Well, I think that this was worked out, to make sure that I came back again. But from the beginning, I had persuaded the University of Chicago to limit this offer to a year and a half. They would have preferred if I had accepted an offer for three years, and if the Copenhagen University hadn't managed to get this professorship, I would have returned anyway, and continued as a lecturer.
Hoddeson:How did you find working in Denmark after your stay in the United States?
Stromgren:Well, in 1938, I still planned for and hoped for active collaboration with Yerkes, even if I didn't go there.
Baym:This would be by mail?
Stromgren:Yes, or by visits. But that of course was all interrupted in 1939. Nothing came of it, until after the war. Struve came to Europe and to Copenhagen in 1946, and he invited me to come as a visiting professor for the period from 1947 to 1948, some ten months. So at that time, we just took up the connections again. But a lot of time had been lost, and during the war years, of course it was rather limited, what I could do, when you think of the plans we had then. They just had to be postponed.
Hoddeson:Before we get into the work in Denmark under the Occupation, I'd like to ask one final question about the period before the war in the United States, in Chicago, about the European astronomers who worked in the United States. What influence do you think the emigres had?
Yes. They did influence the development of astronomy in the States. But not to the extent that the influx of physicists influenced particularly theoretical physics. The situation in the thirties was largely that Europe — European astronomers, in a way, left big telescope astronomy to the Americans. But they felt that in theory, they were ahead, and as we were talking about last time, there wasn't too much (in the States].
This changed, during the thirties, in that — well, at Yerkes, it was the attempt in connection with inviting Kuiper, Chandrasekhar, who, in a way, had a European background, and myself. At Harvard, Shapley tried to persuade Rosseland to come. He didn't succeed, but Bok came. And later on, Martin Schwartzchild, and there were a number of others who, with their European background, influenced the development, particularly perhaps in connection with the building up of graduate instruction, of which there was very little.
There was a Center at Princeton, simply Henry Norris Russell and those who worked with him, and there were beginnings at Harvard and Berkeley, but all told, it wasn't very much. Although it isn't as spectacular as what had happened around the turn of the century in mathematics, and in the thirties in physics, the European influence must have played a role. I would definitely say that.
Baym:Why did the Europeans leave big telescope astronomy to the United States?
Well, it's difficult to say, but clearly, it had to do with the climatic conditions. And with the decision not to engage in observation astronomy in remote places.
I haven't mentioned this, I think, but in the early thirties one of the younger German astronomers with whom I had a number of discussions was Ludwig Biermann, and we still have close contacts, and he knew the development in those days in Germany quite well. He was then at Berlin-Babelsberg and he was convinced that the decision which led, as far as Germany is concerned, to this situation was taken when it was decided to move the Berlin Observatory to a suburb, Berlin-Babelsberg.
Well, it seemed far from Berlin, but still it was not a good climate, and the emphasis was on astronomy, and the big refractor for observations of, for instance, satellites, important work — a meridian circle transit instrument, and vertical circle, for astrometry, and then one reflector, with spectrographic equipment, but very limited, and in that climate, the output never became very important.
Now, if at that time, he says, it had been decided to go say to the Canary Islands, where experience — there was some experience from eclipse expeditions — then, maybe it would have looked differently. On the other hand, if the war, the first World War, had come immediately after these efforts had started, it may not have come to anything. Another possibility was German Southwest Africa, in the neighborhood of [ ]. But it must have been really a formidable problem, to operate an observatory at such distance.
Still, both the British and the Dutch later on made attempts, although rather limited attempts, in that direction.
Baym:The University of Michigan had an astronomical observatory in Africa.
Stromgren:Yes, and so had Harvard, in South Africa, Harvard [ ] place. But as for the French — their main effort came just before World War II, when they founded at the same time the Institut Astrophysique, and particularly the Observatoire de Haute of Provence France was an important development. Although because of the war, it was only in the fifties that it was fully developed, at the time when there was an explosive development in America, so that Europe was still behind. Although the Haute Provence was fairly good observing, it's not good enough. A change came really only, I would say, during Baade's stay as a visiting professor in Leyden.
Stromgren:In the early fifties — when he and Oort discussed the situation, and reached the conclusion that this could only be remedied through collaboration of a number of the countries in Europe, and that led then to the European Southern Observatory Association, ESO. Well, it was a blow that the British decided not to join, so that the beginning it was France, Germany, Holland, Belgium and Sweden, and then Denmark joined in 1967. And this has really changed the picture, even without Britain. The British decided to wait, and the result was the Anglo-Australian Observatory and today, the effort to build a Northern Hemisphere Observatory, to move the Isaac Newton telescope there, and to build perhaps a four-meter telescope, or even somewhat larger. But even after it had been decided to unite and build this European Southern Observatory, the start was quite slow. One could say that the change came really not in connection with ESO, but with the efforts in radio astronomy, particularly in Holland and in England. This started right after the war, and there was a period when, in observational astronomy in that field, Europe was even a little ahead of America. And there was a healthy balance.
Baym:Were you aware of the possibilities of radio astronomy in the thirties?
Stromgren:Well, I had some contacts with people who were pioneering in those days.
Stromgren:No, it wasn't correspondence. It was on the campus. I never had any direct contact with Jansky. For the moment the name escapes me —
Stromgren:Yes. Precisely. And right after the war, let's see, Reber came — a course was given on the campus on interstellar matter. I saw him there. And right after the war, when I was visiting Princeton, Reber took me to see the radio telescope that he had built in his back yard. But because of the war, I was not in contact with those who looked at the problems further. For instance, I only heard about van de Hulst's work right after the war. Then it was immediately clear that there was a big future. And when I left for the States in 1950, that year, I remember very clearly, there was in Copenhagen a visit by Bolton, and it was very impressive, to hear what he thought of the future of radio astronomy. I had hoped when I went back to Yerkes in 1950 that we could go into that field, based on McDonald, and maybe invite Bolton to come. Now, all that came only somewhat later, and not through Yerkes-McDonald, but when Bolton was invited to join Cal Tech. There was some resistance from the establishment. And for one, Baade in the beginning wasn't convinced that this would be a field that would continue to give exciting results. But in Europe, well everyone else, Oort, saw the possibilities, and of course the groups in England developed rapidly.
Hoddeson:Let's go back to Denmark during the Occupation.
Stromgren:Yes. When the war started in 1939, I had just completed this paper on interstellar matter, and during the war, I didn't do much work on that, since I was cut off from any possibilities of learning about the observational results. We continued here, however, our work on stellar interiors and stellar evolution. During the war, Anders Reiz whom you know came from Lund. He was allowed to go forth and back, and we worked together on a problem concerning stellar evolution. It wasn't trying to compute tracks, but just to see if you could understand the structure of the giant star. It was the paradox that Eddington had formulated in INTERNAL CONSTITUTION OF THE STARS. When you compare Capella and the sun, you use the standard model, then you have a much higher luminosity with a much lower density, and much lower temperature. So, one very important development was the one that was suggested by Gamow, that what happens in giant stars is that outside the inert core you have a shell, and there the temperature might be high enough that you have both conditions fulfilled — you have hydrogen fuel and you have sufficient temperature. So this is what Reiz and I tried to do, just again integrate from the outside, and go through a shell that has sufficiently high temperature, and then finish it of with this inert helium core, and that worked. That is the model that had the right temperature for the shell. So it was a step forward, because Gamow's ideas were only qualitative, and we saw that this was probably one viable model for a giant star. But it was still very far from understanding the evolution, because we didn't go step by step. And really, this only became possible when you could do it with the electronic computer.
Hoddeson:Did you know Gamow?
Stromgren:Yes, I knew him very well, from the early days at the Bohr Institute.
Hoddeson:Did you discuss the application of his ideas?
Stromgren:No. This I had only heard of, and we did not meet, then, and he had published this idea.
Hoddeson:By then you had assumed directorship of the Copenhagen Observatory.
Yes. I was director here of a very small observatory, and my administrative work all had to do with preparations for the build-up that came later. There were no observational facilities, and I had proposed, even before the war, that the university build an observatory west of Copenhagen, in the best conditions that you can have in Denmark. I felt that even if we could have access to big instruments, being invited as visitors, you must have a firm basis and you must create a tradition, and for that it was necessary to have even a small observatory. And that was the aim.
We started out during the war looking for the sites. If you know conditions you can eliminate lots of areas, and in the end there were two sites. We had small telescopes in both and investigated seeing. This was done by a method that was for the first time tried on a large scale, where you have a telescope, you cover the aperture, with two openings that are sufficiently far apart that the turbulence in the atmosphere is not correlated. Then you have a prism, so that you can form two images, and the disturbance is not correlated.
Now, Isaac Newton knew that when you have a small aperture, the image is sharp, but it moves around, and you then measure the relative displacements of these images, and you record images of very bright stars, at a time that must be as short as 1/100 of a second, because that was the time scale for these disturbances. We proved then that it is indeed the time scale, that there is correlation between the locations, at intervals of a few hundredths, but no correlations at a few tenths, so that is the kind of scale.
And we did this work both in Copenhagen and in the place that was later adopted for the observatory. I once told the administration at the university that in the process we had measured 16,000 star images. This was the number. But he must have misunderstood, because he was overheard to tell other people that we had investigated 16,000 sites. (laughter)
Baym:Which was the other site that you had considered?
Stromgren:Near Elsinore there is a state forest, and in some ways it seemed good, but there was the danger that the light from Copenhagen would be too much, and I'm glad we didn't choose it. It was rather clear that it was the other.
Baym:Did you have contacts with Lindblad during the war?
Stromgren:Yes. I was allowed to go to Stockholm and spend two weeks there. I gave some lectures. And we had also contacts through the war, by correspondence. I knew how his work was going, which was essentially the theoretical work on galactic dynamics.
Hoddeson:Generally, were you able to follow astrophysics and astronomy in other countries during the war?
Stromgren:Well, some things, we only learned about after the war, but there was, via Sweden, some exchange of literature.
Baym:Had you learned about Baade's work on populations?
Stromgren:Yes. We heard about that. Now, this was part of the activities, to prepare for building the new observatory, and I did take up some problems that I had thought of before in optics in that connection. I had, in the thirties, written an article that gave the theory of the Schmidt telescope, which was quite new at the time. Then, I continued, and wrote the paper on general ray tracing — you may have seen that. The advantage of working on that was that on days when everything looked dark, it was routine work. The optical sine papers were based on calculation of sines, with ten figures, and rounding off, purely routine, which was right on such days. They had wide circulation and use, in the years immediately after the war, before the advent of the electronic calculator. Still, there may be people sitting, working by themselves, who can use that sort of table. Another thing was on a lens system, the equivalent of the Schmidt telescope. You know, if you have a Zeiss and a Meer, and you use a Schmidt correcting plate, you can make an aplantatic system. I asked myself a question: if you have a Fraunhofer lens and are willing to modify the curvature of the surfaces, and to introduce in the middle of the telescope a Schmidt plate, what can you do? And the result is, you can get very good correction.
Baym:This was theoretical.
Stromgren:Yes. There was an American calculation, and after the war, we had one made, of this type. These were just side activities, also partly in preparation for the development of the observatory. Our main work was now theoretical work on model atmospheres, and this was just follow-up to the work that I did immediately before the war, after the breakthrough in H-, when I had computed model atmospheres for the sun and recalculated the abundances. And in that paper, the conclusion was that if you assume two things — a ratio of hydrogen to all the metals (by metals we mean here, those that yield free electrons) of 8000, that is one thing you assume — and that the abundances of the metals, among themselves, are as have been derived from the meteorites by Goudsmidt. That was fortunate that I chose Goudsmidt's work, and there I was influenced by what Zacharias had told me in Chicago — he knew that work and knew it was the best. Now, if you make those two assumptions, then my conclusion was, you get remarkably close reproduction of the observed spectrum through model atmospheres, because you now do it right with H-. And the abundances followed — when I use this and the Goudsmidt numbers, these are the abundances.
Baym:These are your original numbers?
Yes. In the meantime, remember, I had taken the swing by a factor of ten down and up again. You see, the situation then, to return to your question of the use in stellar interiors of atmospheric parameters, was that there were no reliable calculations for such key elements, but only indications of what 0, N and C were, and they came as we were talking about last time, from Struve's and Swing's work on molecules. But it all indicated a Z of a few hundredths. So that for the first time, if you believe the helium, you had the same ballpark, in the atmosphere and the interior, and that led then, in the fifties, to the approach that you really trust the Z you get from the atmosphere. Now you have what you need, not only opacities and the mean molecular weights but also the energy production, and you can then determine the remaining chemical parameters.
The Astrophysics Volume of the HANDBUCH DER EXPERIMENTAL-PHYSIK, Stromgren’s article is on “Objective Photometrische Methoden.”
ERGEBNISSE DER EXACTEN NATURWISSENSCHAFTEN, 16 (1937) 465.
THE ANNALS OF THE NEW YORK ACADEMY OF SCIENCE, (198), 245–54, August 25, 1972, International Conference on Education and History of Modern Astronomy.
Otto Struve, and Velta Zebergs, ASTRONOMY OF THE 20TH CENTURY, (MacMillan, 1962).
Session I | Session II