Alan H. Guth on James E. Peebles:

Excerpts from: The Inflationary Universe: The Quest for a New Theory of Cosmic Origins. Reading, Massachusetts: Perseus Books (Alan H. Guth, 1997.).

In the early 1960s, Robert H. Dicke spearheaded a very active research group in gravitational physics at Princeton University. The group would get together once a week for an evening seminar or discussion, after which they would usually retire to a nearby restaurant for beer and pizza. These meetings were sometimes held on Friday nights, despite the objection of outraged spouses. One of the participants in that group, P. James E. Peebles, tells the story [1] of a very hot night in the summer of 1964 when the group met in the “ridiculously hot” attic of the Palmer Laboratory with a smaller-than-average number of people.
Dicke described to the group some thoughts he had on the history of the universe. Dicke had decided sometime earlier that he could not accept the idea that all the matter in the universe was created in one flash of a big bang. He preferred an alternative idea that was frequently discussed at the time: Perhaps the universe is oscillating, with successive phases of expansion, contraction, and then expansion again. Dicke was aware, however, that astronomical evidence showed that the earliest stars had condensed from hydrogen and helium only. Elements heavier than helium are synthesized in stars, but they were not present, at least not in significant amounts, when the earliest stars were formed. What, then, has happened to the heavy elements that were synthesized in the previous cycle of cosmic existence? Dicke could invent only one possible answer: At the time of the bounce, when the contraction ended and the expansion began, the universe must have been so hot that the nuclei of heavy elements were shattered into their constituent protons and neutrons. In this way the heavy-element debris from the previous cycle can be cleaned up, so each cycle of the universe can start off fresh. If that were true, then there must be a radiation background – an afterglow of the intense heat – that would continue to permeate the universe.
Although Dicke’s idea of an oscillating universe is no longer favored, the prediction of a radiation background remains relevant. An oscillating universe can be viewed as a succession of big bangs, each of which is nearly indistinguishable from a single-episode big bang. If each bang of the oscillating theory would produce a radiation background, then the unique bang of the big bang theory would have the same effect.
Dicke thought it would be fun to look for the background radiation, so he persuaded two young researchers, Peter G. Roll and David T. Wilkinson, to set up the experiment. Then he turned to Peebles and said, “Why don’t you go an think about the theoretical consequences?” Soon Roll and Wilkinson were busily setting up an antenna on the roof of the geology building. The experiment used a Dicke radiometer tuned to microwaves with a wavelength of 3.2 centimeters. The system included a liquid helium cold load similar to the Crawford Hill experiment, but the horn itself was much smaller – only a foot across.
Following Dicke’s suggestion, Peebles enthusiastically worked out the consequences of a hot early phase in the history of the universe. By early 1965 Peebles had written an article on the cosmic radiation. He found, as Dicke had expected, that the universe today should be uniformly bathed with a background of electromagnetic radiation which is a remnant from the big bang. He was also able to predict the spectrum of the radiation – the way in which the energy density varies with the frequency or wavelength. It would be what physicists call a thermal, or blackbody spectrum, at a temperature of 10 degrees centigrade above absolute zero. (pp. 62–64)

Now we return to the activities of Jim Peebles, who submitted his paper on the theory of the cosmic background radiation to the Physical Review in March 1965. The paper was rejected by the journal, as was a revised version that Peebles wrote in response to the referee report on the first submission. The rejections apparently hinged on the issue of credit given to prior work, particularly the work of Gamow and his collaborators [2]. Letters went back and forth, but agreement was never reached and the paper was never published.
Peebles, however, was very excited about the project. Even before the first submission, Peebles accepted an invitation from Johns Hopkins University, in Baltimore, Maryland, to give a colloquium about the work. He presented the colloquium on February 19, 1965 (he still has the notes!), and the subsequent events represent a spectacular success story of the informal communication networks that exists in the scientific community. In the audience at Peebles’ talk at Johns Hopkins was a radio astronomer from the Carnegie Institution of Washington, D.C., Kenneth Turner. Turner had been an old friend of Peebles’ from graduate student days at Princeton, and in fact Peebles and his wife and two young daughters were all staying with the Turners during their visit. Fascinated by Peebles’ prediction of a background radiation filling the universe, Turner mentioned the colloquium to a fellow radio astronomer, Bernard Burke, also at the Carnegie Institution (although now at MIT). Arno Penzias happened to be a friend of Burke’s, and at the end of a phone conservation about other matters, Burke asked Penzias how the measurements with the Crawford Hill horn were coming. Penzias told Burke about the unexplained signals, and Burke suggested that Penzias might learn something very interesting from the group at Princeton.
Dicke received a phone call from Penzias, and the Princeton crew was soon on the road to Crawford Hill. When Dicke and his collaborators saw the results that Penzias and Wilson were obtaining, they were quickly convinced that the Bell Labs team had made the crucial discovery – the echo of the big bang had been found.
The two groups decided to submit separate papers to the Astrophysical Journal, to be published back-to-back. The Bell Labs would describe the observations, while the Princeton Paper would describe the theoretical interpretation. (pp. 66–67)

Penzias and Wilson had measured their mysterious signal at only one wavelength, so they had no way of knowing whether the spectrum in any way resembled that of a blackbody. If they assumed, however, that it was a blackbody spectrum, then the intensity measured at one wavelength was sufficient to determine the temperature, which was found to be 3.5ºK; more precisely, Penzias and Wilson concluded that it was higher than 2.5ºK and lower than 4.5ºK. (In a recalibration carried out shortly afterward, they reduced their estimate to 3.1ºK, again with an uncertainty of 1ºK in either direction.)
Although the calculation by Peebles had predicted a background temperature of 10ºK, the Princeton physicists were not shaken by the discrepancy. The calculation depended on a number of estimations, so the possibility for error was significant. One of the uncertain numbers needed for the calculation is the mass density of the universe. Peebles had used the value of 7*10^-31 grams per cubic centimeter, a value that had been estimated in 1958 by Jan H. Oort for the mass density in ordinary galaxies. This number is somewhere between 4% and 15% of the critical density that would be needed to close the universe (as discussed in Chapter 3). To change the prediction from 10ºK to 3.5ºK, one would have to assume that the mass density was about twenty times lower than this. The Princeton team concluded, however, that the Oort estimate was probably not reliable enough to rule out such a low-density universe.
While a discrepancy by a factor of 3 in the temperature or a factor of 20 in the mass density sounds very significant, one must remember that basic cosmological numbers such as the Hubble constant or the mass density are very difficult to measure. The article in The New York Times described the 10ºK prediction, and then added without further comment that the 3.5ºK measurement “was considered quite close to the prediction.” When I asked Jim Peebles if the discrepancy in temperatures or mass densities was secretly a cause for worry at the time, he replied:
I remember gathering from a very nice review article by Oort (in a Solvay conference) that there was considerable uncertainty in the measured mean mass density. In short, I don’t remember being worried about the mass density, but I do now wish I had kept a diary!
David Wilkinson was even more emphatic about the lack of concern:
I wasn’t the least bit worried that Jim’s 10ºK prediction was too high. The whole [theoretical] story seemed very fishy at the time, and I didn’t take primordial nucleosynthesis seriously. It seemed like a second order problem compared to showing that the cosmic microwave radiation existed. Also, I probably didn’t understand it.
Over the next year Peebles refined his calculations, submitting a detailed paper to the Astrophysical Journal. This paper was published, and it established the standard techniques that have been used ever since in papers examining the origin of the cosmic background radiation. Most importantly, the refinements eliminated in the discrepancy between the predicted and observed values of the temperature. The observed value of 3ºK to 3.5ºK for the cosmic background temperature was now found to be consistent with the Oort value for the mass density of the universe! Peebles concluded that even a critical density of matter, which he estimated as 25 times larger than the Oort value, would be just barely consistent. The issues involved in this analysis will be discussed in the next chapter. (pp. 69–70)

While Peebles was not the first to study big-bang nucleosynthesis, he was the first to complete the calculation of helium production with the same level of detail with which Alpher, Follin, and Hermann had begun it. Unlike his predecessors, he knew from the beginning that heavier elements can be synthesized in stars, so he was mot disheartened to learn that element formation stops at helium. Beginning with a calculation of the shifting neutron-proton balance that closely paralleled the work of Apher, Follin, and Herman, Peebles extended the calculation to include the nucleosynthesis reactions shown in Figure 5.4.
Peebles also seems to have been the first person to trace the evolution of the blackbody radiation to the present. (The Gamow group had computed the redshifting of the radiation, but they had not examined the interaction of the radiation with the matter of the universe.) Peebles found that until the radiation temperature fell to 3000ºK, the hydrogen gas was so hot that atoms would not form. The gas remained ionized, meaning that the electrons and protons moved independently through space. Since photons interact strongly with charged particles, especially charged particles with a small mass, the photons were constantly scattered by collisions with the electrons. The frequent collisions assured that the matter and radiation stayed at the same temperature, cooling together as the universe expanded.
After about 300,000 years the universe cooled enough for the ionized gas to convert to neutral atoms. Peebles called this process “recombination”, a term still standard in cosmological literature. The prefix “re-“ always seemed out of place, however, since according to the big bang theory the electrons and protons were combining for the first time ever. I asked Peebles if the term was a vestige of Dicke’s belief in an oscillating universe, but he said no. “Recombination” is the word used by physicists who study ionized gases (also called plasmas) under laboratory conditions, so naturally the name was transported to cosmology.
A gas of electrically neutral atoms is very transparent to photons, so a typical photon in the cosmic background radiation has traveled on a straight line from 300,000 years after the big bang until the present. The blackbody spectrum would be maintained as the radiation redshifted, with the temperature falling as the universe expands. Starlight would not interact with the blackbody radiation; photons would be added at much higher frequencies, but the spectrum in the microwave region would be almost completely unaffected. The cosmic background radiation, therefore, gives us effectively a snapshot of the universe just 300,000 years after the big bang. To appreciate the significance of a time as early as 300,000 years, consider an analogy that was used in a 1967 Scientific American article [17] by Peebles and Wilkinson. Compare our observations of the evolving universe with the view downward from the observation platform of the Empire State Building. Street level corresponds to the instant of the big bang. Updating the numbers from the original, the most distant galaxies discovered so far correspond to a view down to the 10^th floor, and the most distant quasars are at about the 7^th floor. The cosmic background radiation is equivalent to a glimpse of something just half an inch above the street! The calculations of big-bang nucleosynthesis reached a new level of sophistication in 1967, when Robert V. Wagoner, Fowler, and Hoyle [18] wrote an intricate computer code that incorporated 144 reactions involving all nuclei up to and including sodium-23 and magnesium-23. They found agreement with Peebles for helium-4 and deuterium production, and they obtained for the first time predictions for helium-3 and lithium-7 production. Nucleosynthesis remains to this day an important topic in cosmology, with the predictions changing slowly with time as the measurements of the reaction rates are refined. A large part of the recent work on nucleosynthesis has been carried out by David N. Schramm of the University of Chicago, Michael S. Turner of the Fermi National Accelerator Laboratory and the University of Chicago, Gary Steigman of Ohio State University, and several other collaborators. (pp. 100–101)