Donald Osterbrock played a leading role in the study of AGNs. He is author of the influential book, Astrophysics of Gaseous Nebulae and Active Galactic Nuclei (1989), and coauthor, with Gary J. Ferland, of the greatly expanded second edition (Osterbrock 1989; Osterbrock and Ferland 2006). At the time of his death in January 2007 he was Professor Emeritus of Astronomy and Astrophysics at the University of California, Santa Cruz.
I have never done any research in cosmology, but as an onlooker I have been interested in it for many, many years. I was inclined toward science from boyhood, partly no doubt because of my father's background as an engineering professor, and my mother's as a chemistry assistant in an industrial laboratory in Cincinnati, where both of them, my brother, and I were all born and grew up. My high school had an excellent library, and in it, and also in books from our local public library, I read a lot about astronomy. I had a small amateur-made reflecting telescope with an alt-azimuth pipe mounting, and could look at the poor images that it produced of the Moon, planets, and bright nebulae like the Ring and Orion. My father took me to occasional meetings of the local amateur astronomical society when a famous professional came to town, and I remember especially Harlow Shapley and Otto Struve.
I graduated from high school six months after Pearl Harbor, and in another seven months I was in the Air Force, training to be a weather observer. On a troopship from Honolulu to Okinawa, we proceeded by way of Eniwetok, Guam, and Saipan, and on the way I first saw the star Foma-lhaut and the Southern Cross. After the war ended, I was able to enter the University of Chicago under the so-called GI Bill of Rights, and in three years completed a bachelor degree in physics, and a master in astronomy and astrophysics. Chicago had the best faculty in physics and astronomy in the country at that time, in my opinion, and I was especially inspired by courses in quantum mechanics, taught by Gregor Wentzel, and nuclear physics, by Enrico Fermi. There were no active cosmologists there, but I attended col-loquia by George Gamow, on what we call the "big bang" today, but he called the "ylem-theory" then, and by Maria Goeppert-Mayer on her new interpretation of the so-called "magic-number" nuclei in terms of nuclear shell structure with strong spin-orbit and spin-spin coupling. These two colloquia seemed quite reasonable to me and, I noticed, to nearly all the professors who were there too.
Then, for three years at Yerkes Observatory I again had excellent teachers, especially Struve, S. (Chandra) Chandrasekhar, W. W. Morgan, Bengt Stromgren and Gerard P. Kuiper. All of them taught us about stars, nebulae, and galaxies, even Kuiper, although he also lectured on the Solar System, on which he had begun working during and just after the war. I did my thesis with Chandra, on the gravitational interaction between stars and cloudy interstellar matter, which we would call "giant molecular clouds" today. For two or three weeks in the summer of 1951, I went to a "summer school," organized by Leo Goldberg at the University of Michigan. I wanted to hear the lectures of George C. McVittie, on hydrodynamics of interstellar clouds (though Chandra advised me not to go - he said he had already taught me more than McVittie knew on the subject!). The "school" was held in the old Detroit Observatory building at the UM campus, where all the professors and grad students had offices. I believe I was the only student from outside UM who attended, and so I shared an office with McVittie and with David Layzer, who had just joined the faculty there that summer, with his fresh Harvard PhD degree.
Once a day we all got together in the main room of the observatory, to have coffee and talk about astronomy. In those conversations McVittie and two of the older professors, Dean McLaughlin and Freeman Miller, were scathing in their remarks on Fred Hoyle's steady state theory of cosmology, involving continuous creation of matter. McVittie was a classical mathematical cosmologist, and I had soon seen from his lectures that Chandra had been right. He had little if any physical insight, and his criticisms of Hoyle's ideas were ridiculous, I thought. Basically, he said continuous creation just couldn't happen, and McLaughlin and Miller chimed in as his conservative claque.
After I completed my PhD at Yerkes in 1952, I was fortunate to be appointed a postdoc at Princeton for a year. There I worked out the internal structure of red-dwarf stars, which turned out to have deep outer convective zones, but radiative centers with the main energy production by the protonproton reaction. I had learned of the problem in Stromgren's stellar-interiors course at Yerkes, and he encouraged me to follow it up at Princeton. Martin Schwarzschild and his students were working on red-giant stars, and he helped me tremendously in my work. Lyman Spitzer, the head of the astronomy department, asked me to teach the stellar atmospheres graduate course the second semester I was at Princeton, so he could spend full time on his research on deriving energy for peaceful uses from controlled nuclear reactions, called Project Matterhorn at that time. I was glad to teach the course; there were only four grad students in it: Andy Skumanich, Jack Rogerson,
George Field and Leonard Searle. As they all had long and successful careers as research astrophysicists, I can't help thinking that at least I didn't hinder them in this first course I ever taught.
Hoyle came to Princeton that year as a visitor, working with Schwarzschild on the structure and evolution of red-giant stars for two or three months. Fred's office was next to mine, in the quiet rear of the old observatory building, and we often discussed his research and mine. He was extremely hard working, brilliant, and knew a lot of astrophysics. I was impressed by Hoyle, and although he was not doing cosmology there at that time, I still had an open mind on it. We never discussed cosmology, so far as I can remember. Hoyle was all business on red giants there, as I was on red dwarfs, and those were the two subjects we talked about.
After one year at Princeton, I was appointed to the faculty of Caltech's then very new astronomy and astrophysics department, headed by Jesse Greenstein. My wife and I drove west in the summer of 1953, stopping for a month at Ann Arbor for a second astrophysics summer school, again organized by Goldberg. This one was much more successful than the earlier one, with Walter Baade and Gamow the two main lecturers, backed up by Ed Salpeter and Kuiper for shorter series of talks. About 30 grad students, postdocs, and young faculty members were there. I was most interested in learning from Baade, but Gamow's lectures, mostly on his cosmology, were quite good. He was always humorous, but with plenty of good ideas. By that time in his life he was a fairly heavy drinker, but it never seemed to mar his thoughts nor his lectures.
Baade was a fantastically inspiring lecturer, and I was glad indeed to have him and Rudolph Minkowski as my chief mentors in Pasadena. At that time the Caltech and Mt. Wilson (now Carnegie) astronomers shared the 200-in and 100-in telescopes, and I worked largely on nebular spectroscopy, with some forays into emission-line galaxies, but never into cosmology. There were too many interesting things for me to do with objects in our own and nearby galaxies. Hoyle came to Caltech two or three times while I was there, mostly to work with Willy Fowler and Geoff and Margaret Burbidge, who came there on visits, on nucleosynthesis in stars. Fred was a visiting professor for one quarter, lecturing on the same subject, and I sat in on most of his lectures. But I never discussed cosmology with him then, nor heard him discuss it with others around the astrophysics lunch table in the faculty club, except to utter an occasional disparaging remark about the "big bang."
From Caltech I went to the University of Wisconsin in Madison with Art Code, to help him build up a full-size graduate astronomy and astrophysics department there. Again, I continued largely observational research there with our smaller telescope, using its excellent photoelectric scanner which made it highly effective for nebular problems.
Then in 1960-19611 had a Guggenheim Fellowship to go back to Princeton on leave, this time as a visiting fellow at the Princeton Institute for Advanced Study, where Bengt Stromgren had recently become the professor of astrophysics, "the man who got Einstein's office." Among the other visiting fellows then were Anne Underhill, who had worked with Bengt at Yerkes and in Copenhagen, Su-shu Huang, another Yerkes PhD, and Hong-Yee Chiu. We had weekly astronomy lunches with Spitzer and Schwarzschild, and Field and Rogerson, who had come back as assistant professor and research associate, respectively, and others. These were held in a faculty cafeteria upstairs in Firestone Library, not as spacious or well appointed as the IAS dining room that was built later, but still quite a step up from the aluminum-sided diner on Nassau Street where we had gone in 1952-1953.
I think Martin suggested to Rogerson and me that we review the status of the helium abundance in the objects we knew best: the Sun, on which Jack had done a lot of research while a Carnegie postdoctoral fellow at Mt. Wilson, and gaseous nebulae, with which I was familiar. I had seen Rogerson often in his two years in Pasadena, and we were good friends.
The helium abundances in nebulae were simple; we used the measurements of the Orion HII region and several planetary nebulae, made by my first PhD thesis student at Caltech, John Mathis, who had also calculated the relations between line-strength and abundances of helium and hydrogen. These were supplemented by somewhat later theoretical calculations by Mike Seaton. Our results were that the helium to hydrogen ratio was very nearly the same for planetary nebulae (mean value N(He)/N(H) = 0.16), and for the Orion nebulae (N(He)/N(H) = 0.15). They contradicted the idea that the helium content in our Galaxy might have increased with time, from when the stars had formed that were at present in the planetary nebula stage (then estimated as 5 x 109 years ago) to today.
For the Sun we used absorption-line strengths Rogerson had measured for weak [OI] lines in the Solar spectrum to determine the relative abundance of oxygen as a representative of the heavy elements (usually called "metals," an especially poor term for all the elements heavier than helium, in my opinion!) to hydrogen. Then from the relative abundances to oxygen of all those heavy elements, often described in earlier years as the "Russell mixture," but using more recent compilations, we derived the abundance ratio by mass, Z/X = 6.4 x 10"2. In this notation X, Y, and Z represent the fractional abundance, by mass, of hydrogen, helium, and heavy elements (where the helium fraction is Y = 1 - X - Z).
The other relation we used for the Sun was derived from a series of Solar interiors models that Ray Weymann had recently calculated at Princeton under the guidance of Schwarzschild. These new models were then current state-of-the-art, taking into account a shallow outer convection zone, an intermediate, unevolved radiative zone, and a large inner radiative but hydrogen-burning region, in which the results of nuclear processes over 5 x 109 years had affected the variation of hydrogen and helium content with distance from the center. Energy production was mostly but not entirely by the proton-proton reaction, and there was no central convective core. These were the best models then available, but in addition I liked them personally because Ray had been the brightest and best undergraduate student I had taught at Caltech, and also because his models took into account revisions and extensions of my early research on red dwarfs by Nelson Limber, my close friend from Yerkes days. Nelson had also gone on to Princeton as a postdoc after me.
The well-observed Solar radius, luminosity and mass gave X = 0.67, Y = 0.29, Z = 0.04 for the original abundances in the Sun, at its formation 4.5 x 109 years ago. This set of abundances is not quite the same as we had derived for the planetaries and the Orion nebula had given, but well within the estimated error, we believed. In the end the best overall fit we adopted was X = 0.64, Y = 0.32, Z = 0.04, essentially unchanged for the past 5 x 109 years. Our evidence was that that the helium abundance in the Sun is essentially the same as the results mentioned above for planetary nebulae and the Orion nebula based on the very straightforward recombination-line theory for H+ and He+.
Although many of the numerical values have been revised slightly on the basis of better measurements and improved theoretical interpretations of nebular and Solar spectra, our conclusion has remained unchanged. The abundance of helium in our Galaxy, and presumably in other galaxies as well, had changed little from their earliest days. Most of the helium must have been formed in the big bang. Personally, I could have accepted the idea that both helium and hydrogen had been created together in a steady state universe, but evidently Hoyle, Hermann Bondi, and Tommy Gold could not, nor could other later theoretical cosmologists.
Rogerson and I had done our paper because Schwarzschild suggested it at the time. I don't remember why he thought it was important, but I don't think it was for cosmology. Certainly I did not have that idea in my mind back then. I was interested in it chiefly because Martin seemed to me so uncertain about what the helium abundance was in stars near the Sun. He had used various abundances for it in his early stellar interiors and evolution papers with students, postdocs, and visitors at Princeton as collaborators. Looking back now (I didn't realize this at the time), he had even used Y = 0 (no helium at all)! This was heresy to me, as all grad students at Yerkes were indoctrinated from early on with the interpretation of the spectral sequence as basically a temperature sequence in stellar atmospheres, all with the same abundances in them, with luminosity as a secondary criterion, but only a very few minor abundance variations which Morgan, Keenan and Kellman (1943) had noted in bright stars, and Nancy Grace Roman (1950) had found more in somewhat fainter ones in her postdoctoral research. It was evident that helium was much more abundant than anything else except hydrogen from the great strengths of its lines in hot stars, though we didn't know just how abundant it might be. All the astronomers I talked with in 1953-1958 at Mt. Wilson and Palomar Observatories had the same general idea, I believe.
Only Martin did not have it in 1952-1953, and he didn't seem to in 19601961, although maybe he was just pretending, to convince Jack and me to prove it. I now realize that Schwarzschild had calculated those models with Y = 0 to compare with earlier calculations by Hoyle and Lyttleton (1942). The assumption Y = 0 agreed with Hoyle's interpretation of the steady state theory. As I mentioned above, I could have accepted continuous creation of both hydrogen and helium if that fitted observational data. Perhaps by that time, 1961, Hoyle was already semiconvinced that continuous creation was dead because he knew from his contacts with American observational astronomers that Y does not equal zero anywhere in our Galaxy. But I may be wrong, and I do not want to put words into his mouth or in Martin's either!
In addition to Burbidge et al. (1957), three early theoretical papers that I know of had treated the expected helium abundance in our Galaxy as a result of nuclear reactions in stars. Burbidge (1958) estimated its increase with time from the approximately known luminosity of the whole Galaxy, Maarten Schmidt (1959) formulated and calculated an early "closed-box" model, and Mathis (1959) carried out a somewhat less exhaustive one. All three assumed that the initial helium content was zero, and built up gradually with time, as a result of nuclear processing in stars and return of matter to interstellar space from evolved stars, but all three found, in one way or another, that this hypothesis would not work, although they did not put it that directly. None of these authors considered how the heavy-element content might have increased; that was still an unknown process.
When the CMBR was discovered in the 1960s, I readily accepted it as a confirmation of the big bang picture. I believed, and still believe, in following the observational evidence, as long as it was based on sound theoretical interpretations. However, I think it is a great mistake to trust any detailed numerical values, derived from observational measurements, too far. The theory is always too simple to match reality "exactly." For instance, I have heard lectures, and seen cosmological papers, in which values of X and Y derived from nebular spectrophotometry are quoted and used to three significant figures. Observers are often overly optimistic in stating their probable errors, and theorists who use them can be even more so. But in addition all the available calculations of the HI and HeI emission-line intensities that I know are based on simplified model nebulae, either with one "mean" temperature and one "mean" electron density, or on models in which local means, varying only with distance from the photoionizing star or stars, are used. Yet direct images of nebulae show that down to the finest resolution we have been able to achieve to date, even at excellent seeing-sites on high desert mountains or from space with the Hubble Space Telescope, fine structure, "filaments," and "clumps" are present in nebulae. No doubt these contain a range of densities, temperatures, and excitation conditions down to very small scales. The "mean" values may not represent these conditions to high accuracy, as many current papers are showing. As our understanding of the effects of fine structure, and also perhaps of hydromagnetic heating of nebular gas, improves, the precision of the derived relative abundance will also increase.
Was this article helpful?