Bernie Burke is the William A. M. Burden Professor of Astrophysics, Emeritus, at the Massachusetts Institute of Technology.
Let me start out with some personal background. When I was a graduate student at MIT, 1950-1953, working in Woody Strandberg's microwave spectroscopy laboratory, I was exposed to radio astronomy through three routes. Woody had known Martin Ryle when he was posted to TRE Malvern (Telecommunications Research Establishment) during the war, as the Radlab representative. He worked with Martin on countermeasures - he said that the tension had been tremendous, and the radar people at TRE were "burnt out." He had heard about the use of Michelson interferometry by Martin, and of the Lloyd's mirror interferometer at CSIRO Radiophysics, and thought they were an excellent example of using cleverness instead of brute force to do radio astronomy. Woody had known Taffy Bowen; he also had known Hanbury Brown and Richard Twiss. The director of MIT's RLE, where I was working, invited Taffy to come to MIT to give three lectures on radio astronomy: one was about the Sun, one about the Moon, and the third about everything else. I was impressed, and was further impressed in the same year, 1951, when I heard Ed Purcell describe the discovery of the 21-cm hydrogen line at the joint Cambridge Monday-night physics colloquium. Through Woody's contacts, I also did some of my thesis work in Charlie Townes's lab at Columbia. It was a rival lab, but the atmosphere was wonderfully open; Charlie showed a new gadget that he was working on, called a "Maser," and explained how it worked.
As the end of my thesis work approached, I had to find a position, and I interviewed for a job with Harald Friis at Holmdel, the Bell Laboratories field station. He emphasized that they were in the telephone business. Radio astronomy was never mentioned. I met and got to know Bob Dicke, a good friend of Woody's, and knew about his K-band radiometer measurements on the roof of Building 20. The famous picture of Bob holding the "shaggy dog" in front of the radiometer horn was well known, and his derivation of the atmospheric K-band absorption that degraded K-band radar was also well-known. I probably knew about his upper limit to the cosmic background (Dicke et al. 1946), but its future connection to radio astronomy did not make much of an impression at the time.
I tried to obtain a Fulbright fellowship with Martin Ryle, but that did not work; I then found out that Merle Tuve, director of the Department of Terrestrial Magnetism (DTM) of the Carnegie Institution of Washington, was starting a radio astronomy effort. I had met Merle at an MIT summer study on undersea warfare (the Hartwell project) in 1950, so I contacted him at the DTM, received a postdoc offer, and joined the fledgling program in September 1953. Merle had imported Graham Smith from Ryle's group at the Cavendish, and my education in radio astronomy began. Our first big project was the 22-Hz Mills cross, and two years later Ken Franklin and I discovered Jupiter's radio bursts. This continued a long tradition of making a discovery in radio astronomy, but not the discovery that the radio telescope had been designed for. I got to know Grote Reber, a marvelous person to talk with, and gained an appreciation of his ability. Fred Haddock called him "not a scientist, but a scientific pioneer," which captures his maverick quality. Reber's (1958) personal account of his motivation and work is a masterly description of how pioneering science is done. There is a curious historical note that can be added. Edwin Hubble, the founder of modern observational cosmology, was taught in elementary school by Grote Reber's mother!
It should be remembered that the state of astronomy in the 1950s was quite different from today. There was an unresolved discrepancy between the Hubble age of the universe and the age of the Earth. There had been a few identifications of radio sources and the two brightest had just been identified: Cygnus A and Cas A, resolving the fierce controversy that had raged, led by Tommy Gold, who maintained that most were extragalactic, and opposed by Martin Ryle, who maintained that most were in our own Galaxy. This had been followed by the bitter controversy between Ryle, who maintained that the 2C source counts disproved the steady state universe, and Bondi, Gold, and Hoyle, who (quite correctly) maintained that the survey was so flawed that it did no such thing.
Now, back to the DTM. In the fall of 1953, Jesse Greenstein and Merle Tuve were at work, arranging a symposium to be held at the Carnegie headquarters in Washington, with the evident intent of instigating a resurgence of radio astronomy in the USA. Jesse was chairman of the National Science Foundation's (NSF) advisory committee on astronomy, the first such group that the NSF had convened, and he could be sure of close attention from that fledgling organization. Merle was well connected throughout the government, and the two, despite some philosophical differences, had considerable influence in official circles. They gathered together an outstanding group of participants, who assembled at the Carnegie headquarters on P Street under the aegis of Vannevar Bush in January 1954. The group included Lee DuBridge, president of Caltech, Leo Goldberg from the University of Michigan, Ed Purcell and Bart Bok from Harvard, Rudolph Minkowski and Walter Baade from Mt. Wilson, the optical astronomers who identified Cygnus A and Cas A, Bernard Mills from CSIRO and Graham Smith, both of whom had provided the accurate positions of radio sources that were needed for the identifications, John Hagen from the NRL, John Kraus from Ohio State, and Charlie Townes from Columbia, who was in the process of inventing the maser and the laser. There were the cosmologist Fred Hoyle, Bob Dicke and Lyman Spitzer from Princeton, Henk van de Hulst, Taffy Bowen, and many other prominent researchers. Lloyd Berkner, who would play a key role in establishing the NRAO, attended; he was president of AUI (Associated Universities Incorporated, a nonprofit corporation composed of representatives from nine northeastern research universities). John Firor and I, along with several of the young people from the Washington area, were also invited. For me, it was a grand introduction to the bright lights of physics and astronomy, and a broad-ranging tutorial in astronomy. The real purpose of the tutorials, however, was aimed at the NSF and Department of Defense officials who attended. Here was a new field of science, clearly related to various national interests, demanding attention from those who were funding science in the USA.
Although radio astronomy was being pursued at the NRL, Ohio State, Cornell, as well as at the DTM plus a fledgling group at Harvard, it was at the Washington Conference that American radio astronomy moved to join Britain and Australia as a major power in radio astronomy. Things moved fast. Caltech, Berkeley, Michigan, Illinois, and Stanford all began major projects. The NRAO was established, and while the 140-ft telescope project writhed in agony, the 300-ft telescope was started as a stopgap measure. The result was that less than two years later, observations began with the 300-ft transit telescope, which was built, as John Findlay put, for the price of sugar - 68 cents per pound. I was an early user, and made a map of the entire visible sky at 234 MHz, including an absolute brightness calibration. The cosmic component had no place in my thinking, for I was pursuing the question of the galactic radio halo, which was much more flattened than the Cavendish measurements implied. The results were published in the Carnegie Year Book, but the map itself was never published. Otherwise, the early 1960s were an eventful time. Otto Struve left the NRAO, an unfortunate case of capping an outstanding career with a conspicuous failure, and he died shortly afterward. Joe Pawsey, from CSIRO Radiophysics, agreed to take his place, but he was stricken by a fatal brain tumor and never took office. In this critical time for the NRAO, David Heeschen was named interim director; in fact, he had been the intellectual leader of the observatory from the beginning, even though he did not have the authority to influence major policy issues such as the finishing of the 140-ft telescope. Meanwhile, the search for a permanent director continued, unsuccessfully, and Heeschen was appointed director in 1962, making official what had been, in fact, the case since the NRAO was founded.
A new direction in radio astronomy was developing at the Bell Labs Holmdel station. The director was now Rudy Kompfner, a physicist with a broad range of vision. As in the case of Karl Jansky, the project started as a system to help plan for telecommunications. The Bell Labs knew that they had to look into satellite communications, and were performing scatter experiments on the Echo satellite, a simple aluminum-foil sphere. For sound engineering reasons, they wanted to develop the best possible low-noise antenna/receiver system, and calibrate it carefully. The frequency was 2390 MHz; the antenna was a shielded horn (it looked like a sugar-scoop), and the low-noise amplifier was a state-of-the art ruby maser. Their results were published by Ohm, the project leader, in the July 1961 Bell System Technical Journal, where they reported that separately the total noise of the system from all components was 18.90±3.00K (Ohm 1961). The total noise, measured on the sky, was 22.2 ± 2.2 K, and this meant that the microwave background of the sky was undetectable. It is said, however, that their initial measurements had smaller error bars, and an implied background temperature of 3.3 K was observed repeatedly, but the engineers talked each other into assigning larger error bars. Whatever the actual facts were, they missed the discovery, and their main fault, as engineers, was that there was an unknown source of noise that they did not pursue. My own contact with this work was almost nil - the case of the dog that did not bark in the night? I would say that I was aware of Ohm's work, but I had not seen the article in BSTJ, and I believe that the discrepancy went unnoticed by my colleagues.
For the young American radio astronomers of that time, it was a marvelous era. We all knew one another, and that included the graduate students. Arno Penzias was a student of Charlie Townes, who, to complete his PhD thesis, had taken his low-noise maser receiver to the NRL and installed it on their 50-ft dish. The entire Washington astronomy community was close-knit, helped by the quarterly community meetings that the Naval Observatory hosted. I knew the graduate students at Caltech, including Bob Wilson. I think that we took it as a good omen when the two decided to go to work at the Bell Labs field station at Holmdel.
Arno Penzias joined the Bell Labs in 1961, and was followed a year later by Bob Wilson. Here I have a personal story to tell. In late 1964 I shared an airplane journey with Arno (my recollection it was to Montreal), and I asked him what his plans were at Holmdel. He said that he was going to determine the absolute brightness of the sky at C-band. I said that he would certainly find it was so low that it was undetectable, based on the synchrotron spectrum, and he said yes, he knew that, but it had never been measured and the equipment at Holmdel was the best in the world for that purpose. Arno remembers that I mentioned the earlier upper limit set by Dicke at the Radlab. I don't remember that explicitly, but since I was aware of the measurement, it is entirely possible that I did.
Two years later, in 1965, a colleague at the DTM, Ken Turner (a PhD from Dicke's lab at Princeton), told me about an interesting colloquium that he had attended at the Johns Hopkins Applied Physics Lab. Jim Peebles, a theorist working in Dicke's group, said that there was good reason to suspect that if the "big bang" cosmology was correct, there should be a remnant microwave glow in the sky, the redshifted remnant of the time when the hot gas recombined at a redshift of about a thousand, and Dicke's group were in the process of measuring it. At the DTM, we had a lunch club, with the staff taking weekly turns as cook. It is my recollection that, on the very same day that Ken told me the news, the telephone rang during lunchtime (it may have been a day later, but the interval was very short). I was called to the telephone, and it was Arno, calling about some side issue, possibly about URSI (Union Radio-Scientifique Internationale) matters, and after we finished our business, I asked Arno "How is that crazy experiment of yours coming?" Arno replied "We have something we don't understand." I then said "You probably should call Bob Dicke at Princeton to discuss it." Arno called Princeton, talked to Bob while he was meeting with his group, and the rest is history. Penzias and Wilson received the Nobel Prize, quite deservedly, but it is a shame that it was not shared with Dicke. He shares the distinction of many friends who might have become Nobel laureates, were well deserving of the honor, but who were passed over.
Another footnote story can be told about how Bell Labs profited in a practical way from the CMBR discovery. Rudy Kompfner told me that the space relay system that Ohm's work had been designed for needed a reliable calibration system for the relay stations in the field. The engineers planned to launch a calibration satellite to do the job, when Arno and Bob pointed out that there were radio sources already in the sky that could calibrate the system, with no cost to Bell Labs! Shortly afterward, the engineers were contemplating an 8-mm telephone relay system, but again a calibrator was needed, especially to get a statistical record of atmospheric attenuation. Again, a calibration satellite was proposed, and again Arno and Bob pointed out that the Sun could serve as the calibrator, again at zero cost! The radio astronomy program saved Bell Labs several hundred million dollars in satellites that were not needed.
Many discoveries have precursors, and the discovery of the CMBR has some history of that sort. Joe Weber at the University of Maryland told me this example. He served in the USA Navy in the Pacific in World War II. After Joe resigned his commission in 1948, his expertise earned the offer of a professorship at the University of Maryland, provided he get a PhD. That led to an interview by George Gamow, who was in the midst of his calculations of a nuclear "big bang" at the beginning of the universe. Gamow asked him "Young man, what do you do?" Joe answered, "I'm a microwave physicist." Gamow replied "I'm sorry, but we don't have anything suitable for you at George Washington." It does not seem likely that Gamow had seriously considered how the relic radiation might be detected.
Gamow may have missed an opportunity, but his two younger associates, Ralph Alpher and Bob Herman, may not have. After Joe Weber told me his story about his interview with Gamow, he continued with a second story that illustrates how major discoveries have antecedents, might-have-beens, that for one reason or another did not happen. This was true for pulsars, and particularly for the Crab pulsar; both the radio and optical discoveries had failed precursors. Joe's story is that Alpher and Herman visited the NRL, almost certainly in 1948 or 1949, to see if detection of the microwave background was a possiblity. In radio astronomy, NRL was the only show in town at that time, and it is likely that they talked to the head of the radio astronomy group, John Hagen. He told them that the experiment was too hard, so they did not pursue the matter. I asked Ed McClain, one of their talented young engineers, if he had ever heard from Hagen about the experiment, and he said that he certainly had not. I remember Ed saying "That's odd, because John had a very good nose for new science." I doubt that Joe Weber's story is incorrect, because the Washington radio community was a close group, where everybody knew everybody, but it is, nevertheless, a second-hand story. Might the NRL group have succeeded? Hagen had a powerful team, including Ed McClain, Connie Mayer, Fred Haddock, and Russell Sloanaker, all of them talented microwave engineers and good physicists, familiar with the latest microwave technology. On the other hand, they had concentrated mostly on Solar radio astronomy, a strong source where sensitivity is less important, but receiver stability is vital. Nevertheless, they were clearly interested in fainter radio sources because, in 1950, they persuaded the Navy to fund a steerable 50-ft dish, which they placed on top of the central building at NRL (a terrible location). A simple calculation shows that they might have been able to do the experiment: their crystal mixers had a double-side-band noise temperature in the range 2000-3000 K. Typically, their IF bandwidth was 10 MHz, so a Dicke radiometer would have had an rms fluctuation of about 1K for a one second integration. Connie Mayer, in particular, was meticulous in calibrating radiometers, and all of them knew about the hazards of atmospheric and stray radiation. They had access to liquid helium, so they could make cold loads. I conclude that they had a good chance of being successful if NRL gave them the resources. The conclusion, though, is clear: the experiment was not pursued, and it joins the long list of lost opportunities in science. Arno Penzias and Bob Wilson, on the other hand, were at the right place, at the right time, and their work is a model of how forefront science should be done.
A brief coda is in order. Kipling's ditty
As the dog returns to its vomit, and the sow returns to its mire, and the burnt fool's bandaged finger goes wobbling back to the fire ...
comes to mind, for after I returned to MIT in 1965, the lure of the CMBR pulled me in. I had a talented graduate student, Marty Ewing, and along with Dave Staelin we hatched a plan to measure the CMBR at a shorter wavelength. We chose 9 mm, because atmospheric transparency is good there, and we chose White Mountain, east of the Owens Valley, as the obvious site. Nello Pace of Berkeley had established a high-altitude physiology lab there, so there were electric power, living facilities, and road access. Common sense said that it had to be, at 12,400-ft altitude and east of the Sierras, an uncommonly good place to do the experiment. The Princeton group evidently thought so too, and so in the summer of 1967, side-by-side, we measured the CMBR. The leaders of the Princeton team were Dave Wilkinson and Bruce Partridge, and we became friendly competitors. They were carrying out their measurements at three wavelengths, a better experimental design, but at least we confirmed their results, using a different calibration technique. We used a Dicke radiometer, switching against a liquid helium cold load, and calibrated the overall system by using a helium-cooled "shaggy dog" (actually, a shaggy egg-crate) (Ewing, Burke and Staelin 1967). Dave Wilkinson's group calibrated by using the same reflector to look at the sky and to look down at an egg-crate in a bath of liquid helium (Stokes, Partridge and Wilkinson 1967).
There is a final twist to the story. In about 1970, there was a rocket experiment that tried to measure the CMBR temperature above the Planck maximum, and they found that the sky was much too hot. I doubted the result (which turned out to be caused, not by a hot universe but by hot rocket gases), and sought out Rainer Weiss, a colleague at MIT and a friend of many years. Rai is a great experimental physicist, at the time doing fancy things with lasers, and I think I communicated my enthusiasm. A balloon-borne radiometer was obviously the way to go. In addition to sending data by a radio link, Rai wanted an on-board recorder, and this I borrowed from my close colleague Al Barrett, who had been carrying out balloon radiometer observations of the Earth and its atmosphere for some time. Al was reluctant to lend his precious gear, and on the first flight the wrong squib was fired, and the experiment fell 100,000 ft to the Earth. Al's recorder was among the casualties. I told Al that Rai would buy him a better recorder, and the next balloon experiments worked: the CMBR still showed 3 K beyond the peak. Otherwise, I had little to do with the experiment (Lyman Page was the graduate student who helped Rai with the heavy lifting) but some years later, Rai paid me the ultimate compliment: "Bernie, you wrecked my lab."
A more extended history is given in my article, Early Years of Radio Astronomy in the U.S., Burke (2005).
Was this article helpful?