George Gamow and Ralph Alpher: a review of their cosmological collaboration as mentor and protégé 1942–1955

Abstract

George Antonovich Gamow (1904–1968) and Ralph Asher Alpher (1921–2007) were associates from 1942 until 1968. In this paper, we examine an intense period of collaboration at George Washington University. Our inquiry pivots on a collection of 53 letters and postcards in the Library of Congress (LoC) that Alpher received from Gamow during his absences from Washington DC. In order to set our examination of the letters in their historical context, we present brief biographies of Gamow and Alpher, summarise the state that nuclear astrophysics had already reached by 1945, and examine the initial impact of the αβγ paper. We conducted detailed analysis of twenty of the LoC letters which documents successive attempts by Alpher and Gamow to address the deficiencies in their model of primordial element building by neutron-capture in the big bang. We give a detailed account of the interactions between Gamow writing from Los Alamos, New Mexico, and his two co-workers Alpher and Robert Herman in Washington DC. The correspondence brings their enthusiasm and commitment to life as they react to the advances and setbacks they encountered. Our narrative illustrates the remarkable partnership that Gamow and Alpher shared, a this was, infused with friendship and therein scientific discovery.

1 Introduction

Ralph Alpher and George Gamow feature prominently in popular histories of modern cosmology for their outstanding contribution to the theory of the origin of the chemical elements by nucleosynthesis in the Big Bang. That was the topic of Alpher’s doctoral thesis which Gamow supervised. After gaining his doctorate, Alpher continued to work on the physical conditions in the earliest phase of the Big Bang in a productive collaboration with Robert Herman (1914–1997) which soon resulted in a predicted temperature of 5 K for the fossil radiation created by the Big Bang. This prediction had no impact on physics or cosmology until the remarkable discovery the microwave background by Arno Penzias and Robert Wilson, announced in May 1965, by which time Alpher had been completely forgotten [1]. This serendipitous discovery set off a paradigm shift in cosmology, for which Penzias and Wilson received a half share of the Nobel Prize in Physics 1978. The 2006 Nobel Prize was awarded to John C. Mather and George F. Smoot “for their discovery of the blackbody form and anisotropy of the cosmic microwave background radiation.” Six decades passed from Alpher’s concept that radiation have dominated the early universe and his prediction of 5 K for the present temperature of the relict radiation to the awarding of the 2006 Nobel Prize.

During the first half of this ~ 60-year time span, many references in the literature unwittingly credited the crucial calculation (5 K) to Gamow solo, or Gamow et al., or Gamow and his students. Although this silent treatment of Alpher’s prediction has been critically documented by Kragh [2, 177–185] [3, 101–141], by Alpher’s son Victor [4] and many others, it has persisted. Consider, for example, the public domain text from the Nobel Foundation for the 2006 Prize for …

… a theory predicting microwave background had already been developed in the 1940s (by Alpher, Gamow and Herman). And the discovery therefore made an important contribution to the ongoing discussions about the origins of the universe [5].

This text exaggerates Gamow’s role because he played no part in elaborating the detailed theory that led to a prediction of the background radiation by Alpher and Herman [3, 6]. Furthermore, an archived document from the Royal Swedish Academy has this correction (emphasis added):

Some 16 years earlier Alpher and Herman (Alpher and Herman 1948, 1949; correction of referenced papers made 2017) had predicted that there should be a relic radiation field penetrating the Universe [7].

The adjustment made in 2017 corrected the misattribution by removing Gamow’s name.

2 Introducing George Gamow

Gamow was a member of a select group of physicists who first developed quantum theory and then applied it to the atomic nucleus to account for radioactivity. In 1921, he enrolled at the mathematics and physics faculty of Novorossia University in Odesa, which he described as “to a large extent non-operative” apart from a strong group of mathematicians [9, 24]. After his first year in Odesa, Gamow switched to Petrograd (now St Petersburg), where he secured a position at the meteorological observatory and formed his own circle of mathematician colleagues. From 1922 to 1928, he immersed himself in postgraduate research in physics in the company of Viktor Ambartsumian, Matvei Bronstein (executed February 1938, victim of Stalin’s Great Purge), Alexander Friedman, Yakov Frenkel, Dimitri Ivanenko, and Lev Landau (arrested 1938, one-year prison sentence for anti-Soviet activity) [8]. Friedman worked tirelessly to raise the international standing of Soviet physics. He made a breakthrough in relativistic cosmology by introducing the possibility that the radius of the universe varied with time [9, 10]. Gamow eagerly followed Friedman’s graduate course on general relativity. Following Friedman’s tragic death from typhoid fever (September 1925), Gamow and his young circle (Bronstein, Ivanenko and Landau) self-educated themselves on the new wave mechanics of Heisenberg and Schrödinger.

In summer 1928, the university awarded him funding for a four-month study visit to Germany and when he entered Max Born’s Institute of Theoretical Physics in Göttingen in June he found it buzzing with “excitement caused by wave and matrix mechanics” [11]. Here Gamow accounted for alpha decay as quantum–mechanical tunnelling through a potential barrier [12]. Gamow had made the first successful application of quantum theory to nuclear phenomena and he stimulated colleagues to develop it. He established an enduring reputation as a nuclear physicist with his Oxford University Press monograph on radioactive decay [13]. In late 1930, Gamow returned to Stalin’s Russia, where “science had now become one of the weapons for fighting the capitalist world” [9, 92]. By 1932, Gamow had decided to abandon his motherland because of the conflict between Soviet philosophy and western science. He was thus taken aback when the Soviets unexpectedly appointed him their national representative to the seventh Solvay Physics Conference in Brussels, October 22–29, 1933. In Brussels, Gamow decided not to return to the Soviet higher education system which suppressed the principles of liberal knowledge and institutional self-governance in the academy [14]. He and his wife Rho embarked the Danish liner S/S United States for New York City in summer 1934 [9, 104–107]. After a few months in the US, Gamow accepted an offer to join the physics department at George Washington University (GWU), subject to two conditions: that his colleague Edward Teller be invited to join the GWU faculty, and an annual international conference of the highest standard on theoretical physics to be under sponsorship of GWU and the Carnegie Institution.

A thematic thread linking the Gamow–Alpher correspondence at the core of this inquiry is Gamow’s early work on nucleosynthesis. Important stepping-stones included: (a) Richard Tolman’s proposal that high temperatures were needed to ensure that equilibrium production of hydrogen and helium would result in the present abundances [15]. (b) in 1929 Atkinson and Houtermans, encouraged by Gamow, investigated barrier penetration in fusion reactions [16]; (c) Houtermans (1931) developed the first nucleosynthesis theory in which elements are built up step by step from lighter ones [17]; (d) the recognition that it was impossible to account for the abundances of the light and heavy elements via equilibrium at a single temperature and density [18]. In 1935, Gamow proposed a model of stellar nucleosynthesis by successive capture of neutrons (with subsequent β-decay) rather than charged particles [19]. Gamow expressed interest “… in the role played by stellar [cores] in the formation of heavy elements”, remarking that “the many different phenomena inside a star” give rise to the formation of different elements. But the optimism slowly dimmed as it emerged that no single theory could account for the abundance data.

At the 1942, Washington conference organised by Gamow and Teller, Chandrasekhar and Henrich, with others, accepted that nucleosynthesis of the heavier elements could not have arisen under conditions of thermodynamic equilibrium in stars. They floated the idea that elements formed at earlier prestellar stage of the universe. With a temperature ~ 10¹⁰ K and density 10⁷ g cm⁻³ the known abundance curve from oxygen to sulphur agreed fairly satisfactorily with the theory, although they cautioned such apparent agreement should not be over-stressed [20]. Encouraged by this scenario, Gamow (1946) proposed that free neutrons had gradually coagulated “into larger and larger nuclear complexes which later turned into the various atomic species by subsequent processes of β-emission” [21, 22]. We continue this story below when Gamow accepted Alpher as his research student.

3 Introducing Ralph Alpher

Ralph Alpher, the son of Russian émigré Jews, was born on 3 February 1921 in Washington DC, where he would receive all of his education. Ralph’s father Samuel Ilfirovich (Ильфиpoвич) had fled Belarusia in 1904 as a teenager, and likewise, his Latvian mother Rose arrived in the USA in her teens. Ralph’s upbringing was in a settled home in Washington DC, where his parents conversed in Yiddish. As an adult, Ralph Alpher worked assiduously for acceptance as a loyal and industrious American citizen regardless of heritage. At Theodore Roosevelt High School, three teachers nurtured his curiosity in astronomy, physics and chemistry. His English teacher guided him to the elegantly written popular books on astronomy by Jeans and Eddington. An imaginative physics teacher with a hands-on approach to group work required classmates to strip down a Model-T Ford, describe the working of each component in terms of physics, and then “put it back together again for the next class”. Alpher’s success in academic studies culminated in a prestigious scholarship at the Massachusetts Institute of Technology, but the offer was withdrawn peremptorily after an exploratory meeting with an alumnus in which he was subjected to intense questioning. Devastated by this sudden revocation, he silently speculated it was on account of his Jewish heritage. Following this rebuff, his main concern was supporting the family’s meagre income after the Depression had wiped out his father’s business. He worked as the stage manager for the theatre at Theodore Roosevelt High School bringing home income that the family sorely needed. Another practical step he took was self-improvement: he obtained his Gregg Shorthand certificate in 1937. His first clerking position was at the National Institute of Standards (now NIST), which introduced him to the importance of precision and clarity in physics research. He also enrolled at GWU for night school classes in chemistry.

From 1938 to 1940, the Department of Terrestrial Magnetism of the Carnegie Institution of Washington engaged Alpher to support Scott E. Forbush (1904–1984). During evenings, he took Edward Teller’s course on physics at GWU finding the lectures so stimulating (and fun) that he switched from chemistry to physics, never looking back. When the federal government assumed a war footing in 1940, the scientific research effort in Washington was reorganised for military objectives. Alpher was hired in 1940 as a physicist at the Naval Ordnance Laboratory, contracted to undertake classified work. He thus became acquainted with many outstanding physicists in the Washington area. He made significant contributions to the development of degaussing technology for countering the magnetic field produced by ship hulls, which led to the retrofitting of naval vessels and the redesign of newly commissioned ships to include degaussing coils. By 1943, Alpher’s focus turned to the detection of submarines with an airborne magnetometer and the development of magnetic proximity exploders (fuzes) for air-launched torpedoes. Throughout this period, he worked in applied physics research at the doctoral level. Ironically he only completed the GWU night school courses for his master’s degree in 1945.

4 The αβγ saga 1942–1948

When interviewed four decades after starting doctoral training under Gamow, Alpher recalled:

I found him a tremendously stimulating person. And he obviously loved physics, and enjoyed it, and he conveyed to me as a student a sense of enthusiasm which was hard to ignore. When it came time to do a master’s work, I asked if I could work with him. He said “sure” and that was delightful [23].

Alpher’s formal doctoral training under Gamow’s supervision commenced in summer 1945. For Gamow, who had no inclination to carry out detailed mathematical analysis, it was an attractive partnership: Alpher was a meticulous and talented applied mathematician, capable of working unsupervised on the solution of complex arrays of differential equations. Initially Alpher felt comfortable with a thesis topic on cosmogony, the growth of perturbations seeding galaxy formation in a relativistically expanding medium. Unfortunately it led to a negative result, but worse was to befall him: the leading Soviet physicist Evgeny Lifshitz published similar results [24]. Dismayed and distraught by this scoop Alpher impulsively destroyed his notes.

For a new thesis topic, Gamow proposed to Alpher an analysis of the physical conditions for cosmic nucleosynthesis. For the initial state, Gamow posited a universe dominated by neutrons rather than radiation [3, 108].^1,2 Alpher already knew the essential nuclear physics from Gamow’s lectures and Bethe’s extensive reviews [25,26,27]. Alpher’s doctoral thesis project was an analysis of the build-up of elements by successive capture of neutrons in the first few minutes of the rapidly expanding primordial universe. In late June 1946, Alpher attended the 275th meeting of the American Physical Society where precise data on the collision cross-sections of fast neutrons captured by a wide range of nuclei had just received clearance for publication [28, 29]. Donald J. Hughes of the Argonne National Laboratory fulfilled Alpher’s request for a copy of these cross-section measurements. Alpher’s computation accounted for the growth and decay of nuclei but ignored the time-dependence of the decay of the neutrons and the evolution of particle densities as the universe expands. The outcome of the computation of nuclear reactions in the first 300 s of expansion broadly matched the observed cosmic abundances of nuclei as then understood and it accounted for the high cosmic abundance of helium.

Gamow was anxious to establish their priority. To that end, he drafted the Alpher–Bethe–Gamow paper (αβγ) summarising the main findings of Alpher’s dissertation in 546 words (plus one figure) [30]. Alpher was not best pleased (to put it mildly) when Gamow disingenuously slipped in Bethe’s name “in absentia” as a fake co-author.^3,4 The opening sentence of αβγ states that primordial nucleosynthesis was a consequence of “a continuous building-up process arrested by a rapid expansion and cooling of the primordial matter”. Gamow alerted The Washington Post that his graduate student Alpher, in conjunction with Gamow “and Dr H. Bethe of Cornell University” had published “science’s latest theory of the creation of the universe”. The Post subeditor bannered: “World Began in 5 Minutes, New Theory” [31]. This media coverage was the warm-up for Alpher’s thesis defence in late May 1948 at which a formal procession of begowned academics presided over an attentive audience of 300—many of them press reporters. One reporter, Watson Davis, instinctively sensed the general public would find Alpher’s model the creation of the chemical elements enticing: The New York Times syndicated his coverage story to the national press [32, 33].

At the date of his thesis defence (May 1948) Alpher and Gamow knew first-stage αβγ model was too simplistic. The two authors of αβγ conceded that the neutron density dependence on time according to the relativistic theory of the expanding universe was wildly inconsistent with the time required for element build-up that fitted the natural abundance curve of the heavy elements. Since Peebles (2009, 2014, 2020) has documented the many technicalities, we shall not repeat them [6, 34, 35]. In the full version of his dissertation published subsequently, Alpher explained that the data of Hughes on capture cross-sections for 1 MeV neutrons was an important piece of evidence that enabled him to derive formulae “[to] approximately represent the relationship between capture cross-sections, atomic weight and neutron energy.” With such formulae, Alpher obtained expressions for the probability per second that “a nucleus of [a given] atomic weight will capture a neutron.” He found that the temperature must have been well above resonance energies and “well below the binding energy per particle in nuclei” for neutron capture to start. He chose the geometric mean of the upper and lower bounds neutron energy of 0.1 MeV (temperature ~ 10⁹ K). In his thesis, he states:

… the density of blackbody radiation at this temperature would have been … 10 gm/cm³. This is many orders of magnitude greater than the density of matter given by the particular cosmological model used. It would appear, then, that radiation was dominant in determining the behavior of the universe in the early stages of its expansion, and the cosmological model which we have used is not correct. An interpretation of the starting time and initial density for the neutron-capture process will therefore require the development of a new cosmological model.

Here we see here the key development that changed everything: the huge transition from a cold universe dominated by matter) to a hot universe with radiation propelling the universal expansion. This was transformative for physical cosmology. Alpher’s summary of his findings concluded modestly by accepting that the theory “is obviously in preliminary form” [36]. We shall see that the quest for “a new cosmological model” is a recurrent theme in the flow of letters to Alpher when Gamow was absent from Washington DC.

5 Gamow’s letters of 1948

We shall now review the wider impact of αβγ, primarily through letters from Gamow to Alpher conserved by the Library of Congress, which we list in Table 1. We found Peebles’ books and articles (2009, 2014, 2020) indispensable guides to aid our readings of Gamow’s letters [6, 38, 39].

Table 1 Catalogue of the Gamow–Alpher letters 1948–1955 referenced in this paper

In the months immediately following publication of αβγ. Alpher (and Gamow) had concerns that their analysis for creating the light elements developed in his thesis was oversimplified. They knew that the absence of long-lived atomic nuclei at atomic mass 5 and 8 prevented element build-up going beyond helium (the mass gap problem). In his thesis, Alpher side-stepped that problem by inferring that nuclear reactions (then unknown) might carry the build-up past the mass gap. To address the model’s other shortcomings, he and Herman reformulated it to take into account radioactive decay of neutrons. The two colleagues carried out these cosmological revisions in their spare time—evenings and weekends devoted to carrying out the integrations [37]. Meanwhile, Gamow, in his impulsive haste, was making careless blunders in his reworking of the model much to the dismay of Alpher and Herman. Nevertheless, by the end of May 1948 Alpher had improved the curves showing reaction rates as a function of atomic weight. Gamow was clearly delighted with Alpher’s revised result and he worked to publicise it. For example, on 6 April 1948, following an instruction from Gamow, Alpher wrote to Oskar Klein (1894–1977) the greatest Swedish theoretical physicist of the twentieth century, to inform him that “we have found a somewhat better fit … to the observed abundance data”. Alpher wrote that nucleosynthesis “started 310 s after the expansion of the primordial matter began”. Gamow appended a note to Alpher’s letter: informing Klein “I will not be in Zurich but will be in England in Sept. & Oct.” and asking, “Are you going to be there”? Because αβγ was a rushed job Gamow sought to ensure their priority for the scenario of element synthesis in an expanding universe by endorsing the improvements undertaken by Alpher.⁵

5.1 Letter from OSU: 16 June 1948

Gamow departed Washington for Ohio State University where he immediately reminded Alpher to send reprints of αβγ to a “very spotty list” of recipients (one was Klein) to which he would “easily add more persons.” He asked Alpher “to brush it up … so we will form a permanent list which Sherley [Shirley Thomas, typist] will use in addressing the envelope”. Gamow’s “very spotty list” was a roll call of the major influencers in quantum mechanics, nuclear physics and relativistic cosmology. He listed 20 former colleagues with addresses in western Europe; a further six at Princeton; followed by three at Mount Wilson Observatory and another three at Caltech; three in Chicago (Marie G. Meyer being the sole female); two at Cordoba Observatory, Argentina; Nishina and Yukawa in Japan; Einstein was the oldest person listed; Heisenberg, Max Laue and von Weizsäcker in Göttingen were Gamow’s colleagues of longest standing. Gamow was highly motivated to bring αβγ to his European network with all due speed. Gamow signed off: “Please send me also the copy of the new curve you have shown me yesterday.”

5.2 Letter from OSU: 30 June 1948

Two weeks later, Gamow approved the published version of Alpher’s “dissertation submitted … in partial fulfilment of the requirements for the degree of Doctor of Philosophy” [41].

I looked through the m.s. and graphs and mailed them to Tate special delivery. You will get the card from Tate acknowledging receipt of the m.s. It is really very very nice that the problem will be put on IBM mashines (sic) but it must be set really good before they will run it through. Probably, the entire n, H, H², H³, He³, H⁴ family. I have my suspicions that more data on the cross-sections of these nuclei are known than is actually published, and maybe we can get these “for the glory of science.” Why don’t you call Jim Pickard and ask him? Or maybe Chas. Critchfield (now in Los Alamos) is the man to ask.

An introductory footnote to the paper (p. 1577) confirms that αβγ paper was a brief preliminary report of the principal result in Alpher’s dissertation; in this definitive version also for Physical Review, G. Gamow is thanked sincerely “for the stimulating aid and advice received … during the course of this study.” Alpher’s famous footnote (p. 1581) is the first written usage in the English language for almost six centuries of the noun ylem, for the primordial hot matter of the universe.⁶ Another footnote (p. 1587) records that Herman had provided the tabulations of an exponential integral relating universal density and time in the adopted cosmological model.

The science content of the letter of 30 June confirms Gamow’s continuing priority to be understanding the formation of galaxies in an expanding universe. Before departing Washington, Gamow had submitted to Physical Review a paper on the cosmogonical conclusions that could be inferred from their improved model of neutron-capture build-up in a universe dominated by radiation [38]. Consequently, he was now keen to improve their cosmological model. In order “to brush up the origin of galaxies” Gamow employs simple calculus to track the rates of change of the densities of neutrons, protons and deuterons; it was not quite “the entire … family” of nucleons but was a useful start. Gamow urges Alpher to think about what “absolute amount of deuterium (or helium) will be the same as the value necessary to give the amount of heavier elements”.

The 30 June letter is a draft of the material that Gamow had incorporated into two papers: “The Origin of the Elements and the Separation of Galaxies”; secondly “The Evolution of the Universe” [39, 42]. Furthermore, we note that this text is the centrepiece of Appendix VI of the Oxford Monograph (1949) by Gamow and Critchfield, an addendum that Gamow rushed to the publisher in July 1948 [40, 334–337].

The final paragraph of this letter informs us of Gamow’s enthusiasm for the potential of “IBM mashines”:

My lectures here are going well, and everything else, are going O.K. and the local IBM-outfitt starts for me the calculations of the motions of stars newly formed in originally gaseous galaxies. It seems that IBM is really going to be helpful from all sides!

The “local IBM-outfitt” (in Columbus, Ohio) possibly supplied an IBM 604 electronic calculator: in 1948 these machines became available for a monthly rental of $645 (about $10,000 in 2023 dollars). They were hundreds of times faster than earlier electromechanical machines and therefore found many applications in applied mathematics and engineering. At this time, Alpher was still having to manage with electromechanical Marchant and Friden calculators.

5.3 Two letters from OSU: 12 July 1948 and one undated

Gamow announced that he would be in Washington 19–20 July for a joint conference of the Applied Physics Laboratory and the Office of Naval Research. He advised “… we probably will have a lot of time to talk about things” and suggests two topics: implementing the distribution of reprints and further work on the origin of elements.

Reprints-sending. We must make a nice long list of people (here and in Europe) to which the reprints of (α,β,γ) and all the following publications will be send.⁷

The second topic proposed for discussion in Washington concerned the proportion of protons that would remain when nucleosynthesis ceased: Gamow advocated Y = 0.5 for the helium mass fraction. In an undated letter, Gamow sent an afterthought: exclaiming “Oh! Oh! Oh!” Gamow admits that he made an error in the n-p capture cross-section, which required him to repeat the integrations. He writes “This will not change of course the order of magnitudes but will look tidier.” However, that was not to be the end of the matter as we shall see. The adoption of 0.5 for the helium mass fraction Y is an example of Gamow’s fertile intuition at work. Gamow and Critchfield (1949) reference Menzel’s estimates of primeval abundance as X = 0.47 (hydrogen, solar atmosphere) and Y = 0.41 (helium, stellar atmospheres) [45, 282–283]. Peebles remarks that although later evidence gave 0.24 for the primeval helium abundance Gamow’s intuitive guess in 1948 was close enough [40, 139].

5.4 Gamow’s final letter from OSU: 30 July 1948

George sent an appeal for hands-on assistance with his forthcoming seminar presentation:

Two weeks from today I am giving a lecture on origin of elem. & galaxies in Los Alamos and need the curves. I think it would be best to make slides, which we all can use when we need them.

I will need:

(1) Huges curves with indication of small cross sections for mystic numbers.

(2) Abundance curve observed with the correlation to mystic numbers.⁸

(3) Your old curves with n_nDt’s

(4) Your new curve with Bob

[Do] you think you can make these slides in a hurry and send here [i.e. OSU]

Geo.

In the above, “Huges” refers to the published data on capture cross-sections by D. J. Hughes (1946) and others [32]. The third slide requested (3) was from Alpher’s dissertation: it compared relative abundances computed by neutron-capture theory with Goldschmidt’s relative abundance data. The parameter n_nDt is the product of the neutron concentration and the elapsed time since the element formation began. The best fit is n_nDt is 0.81 × 10¹⁸ s cm⁻³.

Gamow’s fourth request for “your new curve with Bob”, refers to the reformulation of the neutron capture theory taking into account explicitly the radioactive decay of neutrons; it refers back to the final sentence of Gamow’s letter dated 16 June 1948 (see Sect. 7.2 above). Much of the extensive computation and analysis necessary for the slides was conducted by Alpher, using Herman’s method. Alpher and Herman (1948) submitted their definitive paper including neutron decay to Physical Review in August. Meanwhile Alpher, Herman and Gamow were writing up the effects of proton reactions on reducing the abundances of Li, Be and B subsequent to the short period of element building by neutron capture. Their brief Letter to Physical Review cites Alpher (1948), and Alpher and Herman (1948), as being in press [41].

5.5 Los Alamos, Thursday [… August 1948]

At OSU and subsequently at Los Alamos, Gamow focussed on understanding the formation of galaxies in the expanding universe. In particular, as we have noted, he investigated the conditions (densities of matter and radiation) for an initial all-neutron state to produce a final hydrogen content of X = 0.5. Gamow’s modelling required the necessary time for a 50:50 balance of matter and radiation densities in the expanding universe to be t = 3.9 × 10¹⁵ s (130 million years). “The gravitational condensation” of galaxies commenced. The masses and diameters of the condensations were

$$ \begin{aligned} & M = 2.7 \times 10^{7} \;{\text{solar}}\;{\text{masses}},\;{\text{and}} \\ & D = 13,000\;{\text{light}}\;{\text{years}}, \\ & {\text{temperature}}\;{\text{of}}\;{\text{the}}\;{\text{gas}}\;T = 340\;{\text{K}}. \\ \end{aligned} $$

When Alpher reworked the calculations in Gamow’s paper The Evolution of the Universe submitted to Nature solo—without the support or indeed opportunity to check by himself and Herman, both were shocked at several errors in Gamow’s calculations [44]. By the time the APL/JHU duo had uncovered the goofs it was almost too late: Alpher placed a long-distance call to New Mexico, triggering an intrusive interruption to Gamow’s generous pouring of preprandial martinis. Four hours later (8 p.m.), after mailing corrections to London and Oxford, a dozy Gamow dashed off a quirky letter of heartfelt gratitude to Alpher:

Geeee … your long distance call almost spoiled my cocktail party, but I fixed it up before the first drink! The mix up is due to the fact that I used a in one sence (radiation density = aT⁴) in one place (: expansion formula) and in another sence (flux = aT⁴) in another (definition of a) It is because I used the second sence in original calculation, and didn’t turn it over everywhere to the “first sence” when putting it down for publication. Thus the only change you have to make in your copy of the m.s. is to put c^11/2 instead of c^23/4 in the formula for a. It does not change the numerical results for M and D, and everything remains just dandy! The expressions for M and D are dimensionally correct (if rad. Density =aT⁴) as you can justify by simple check …..

But my thanks are due to you since I mailed two letter to London (Nature) and Oxford (Clar. Press) asking to correct the formula for a putting c^11/2 instead of c^23/4….

5.6 Alpher predicts a present universal temperature of 5 K

With Gamow’s consent, Alpher and Herman submitted a note to Nature listing four corrections to Gamow’s article [46]. One was an oversight in the physics: not taking into account the effects of magnetic moments in calculating the cross-section for the formation of deuterons from collisions of protons and neutrons. When they recalculated the matter density, they found that the necessary conditions for condensations should include the density of radiation in addition to matter. They concluded that the masses of the integrated values of the matter and radiation densities “intersect at a reasonable time ≅ 10⁷ years, and the masses and [diameters] of condensations at this time become, according to Jeans’ criterion …

$$ \begin{aligned} & M = 3.8 \times 10^{7} \;{\text{solar}}\;{\text{masses}},\;{\text{and}} \\ & D = 22,000\;{\text{light}}\;{\text{years}}, \\ & {\text{temperature}}\;{\text{of}}\;{\text{the}}\;{\text{gas}}\;T = 600\;{\text{K}}{\text{''}}. \\ \end{aligned} $$

Famously, Alpher then commented,

The temperature of the gas at the time of condensation was 600° K and the temperature of the universe at the present time is found to be 5° K.

This appearance in print of a prediction of the present universal temperature was not recognised to be of significance until 1965 [42]. Six weeks after its publication, Alpher and Herman submitted their paper to Physical Review adjusting the physical conditions of their cosmological model of nucleosynthesis to take account of universal expansion. It announced satisfactory agreement when the assumed density of matter at the onset of nucleosynthesis was increased by a factor of ~ 100. Alpher again noted that the present temperature of the radiation would be ~ 5 K:

This mean temperature of the universe is to be interpreted as the background temperature which would result from the uniform expansion alone. However, the thermal energy resulting from nuclear energy production in stars would increase this value [43].

Some twenty years later Alpher told Jim Peebles⁹:

We gave very little thought to the question of detection of this blackbody radiation; three years after World War II microwave radio astronomy was pretty much in its infancy. Moreover, after 1953 all of us were working in quite different directions and giving little further thought to these questions.

Kragh [4] and Peebles [5] have considered to the plausibility of Alpher’s prediction of a temperature of ~ 5 K, given its sensitivity to the choice of unknown initial conditions such as the average density of matter in the early universe, and the practicality of detecting it in 1948–1953. These were serious problems. Gamow took little interest in the prediction, his principal priority at Los Alamos being the conditions for the condensation of galaxies in the early universe. Section IV of Alpher and Herman’s [43] paper discussed Gamow’s notion that “galactic formation occurred at the time when the densities of matter and radiation were equal” [48]. As touched on in 5.7 above, they concluded that condensations formed in congruence with Jeans’ criterion for gravitational instability would lead to “the separation between galaxies at the present time of about 10⁶ light years [~ 0.3 Mpc]”, in general agreement with observations [44]. Nevertheless, the following paragraph seriously questioned the validity of Jeans’ instability “since it does not contain the possible effects of universal expansion, radiation, [general] relativity and low matter density.”

Alpher and Herman hint in conclusion that they have tired of collaborating any further with Gamow on galaxy formation:

Until such time as a physically satisfactory criterion for the formation of galaxies is found, it does not appear to be profitable to delve further into such questions as the variation in galactic mass and size with time of formation.

Kragh succinctly summarises the stage that had been reached:

At the end of 1948, Big Bang cosmology had experienced a drastic development and had now, if only still in outline, turned into a proper theory with quantitative estimates of how the universe had evolved with time. The justification of the new picture of the hot, radiation filled early universe was primarily its agreement with the empirical abundance curve.

Primordial Big-Bang nucleosynthesis was Gamow’s big idea and it was Alpher, aided by his colleague Herman, who developed the idea into a sophisticated scientific theory over the five-year period 1948–1953. Gamow, a self-confessed “very poor mathematician” did not have the time, patience or motivation to carry out the hundreds of hours of manual computations that might lead to no more than a heap of paper in the trash bin.

5.7 Postcard from Pasadena, California 26 November 1948

Gamow sent a brief message to “Alpher & Herman” after talking with Robert W. Christy (1916–2012) at Caltech on incorporating the effect of the expansion of the universe in the neutron-capture theory. Gamow gave the general relativistic equation for time dependency of the scale factor and advises “thus one must integrate slightly differ[ently]”. Alpher and Herman soon commenced this relativistic revision, as we shall see.

5.8 Alpher invited to co-author an astrophysics monograph: 7 January 1949

Gamow’s next request for assistance began with, “I have mailed to you some time ago 100 Universe reprints. Keep them until your and Bob’s Nature reprints come in, and then send them to the list.” Gamow then introduced “Another piece of business”. Oxford University Press had sounded him out about writing a book on the structure and evolution of stars for their acclaimed Monographs on Physics. For peripatetic Gamow, who had become increasingly dependent on writing to balance the books of his personal finances, the Oxford invitation was welcome, but he could not undertake it as the sole author. In 1946, he had brought in his former student Charles Critchfield to recast Structure of Atomic Nucleus and Nuclear Energy Sources. Gamow foresaw that this new challenge for an astrophysics monograph would once again depend on collaboration with a former student. Alpher expressed his willingness to collaborate, but only if Herman too joined as equal partner.

5.9 Gamow vetoes Herman as third author: 12 January 1949

Gamow was unmoved by Alpher’s request.

Just received your letter. I am glad you are interested in co-operating on the book on Stellar Structure and Evolution. I have thought myself about including Bob in this work, but I am afraid it is not a very good idea. As they say: “Three is the crowd,” and the three-author book will become heterogeneous unless different authors are explicitly responsible for different chapters, which does not apply in this case. … I already had some difficulty with Oxford press arranging Critchfield as the co-author of the nuclear book.

But I do not see any reason why two of us couldn’t do it easily in a year or so. You have collected a lot of material for your Master Thesis and it can all be used in somewhat more digested and reviewed form. I do not think that much more than one chapter should be devoted to the problem of the origin of the elements and its cosmological consequences. The thing is developing so that most of the material will be out of date when the book is printed. For example, Fermi and Turkevitch have recently spent quite a lot of time on it and have apparently obtained very satisfactory results (with the exception that they get too much Helium). Turkevitch will be here by the end of the month and I expect to learn a lot from him. I will also see Fermi in Chicago on the way home.

The register of this letter confirms Gamow’s growing scepticism about the mission of Alpher and Herman to further improve the model of nucleosynthesis, which now looked like “last year’s model”, as one might have said in Detroit. Gamow’s parenthetical caveat “with the exception that they get too much helium” refers to the mass gap problem.

6 Origin of light elements: bridging the mass gap

6.1 Letter from Los Alamos: 29 January 1949

The αβγ scheme of element synthesis by successive neutron capture in the expanding universe faced many challenges by 1949. Alpher and Herman continued on their chosen path: nucleosynthesis in the very early universe. Gamow’s letter of 29 January 1949 captures his continuing enthusiasm for cosmogony, in particular the primordial condensation of galaxies in Big-Bang cosmology. Gamow and Teller began research on the origin of galaxies in late 1938, only to be interrupted by war [45]. Gamow’s letter informs Alpher that Nicholas Metropolis and Stanislav Ulam at Los Alamos are joining with Gamow and Teller for “a vigorous attack on the problem of gravitational condensations in the expanding space in the presence of radiation”. The results are “interesting and I should say revolutionary”. The presence of radiation (which had been ignored in the Lifshitz scoop paper of 1946) “plays quite an important role in the formation process”. Alpher and Herman were at this date also aware of the need to include radiation density (5.7 above). Gamow’s consortium had noticed a slight temperature difference between matter and radiation, acting like additional gravitation. In the letter, Gamow suggests it was the agency that triggered condensation:

Radiation plays quite an important role in the formation process. In fact, there exists a gradient of radiation pressure due to the lag between the gas- and radiation temperatures in the expanding space. This produces “radiational instability” which can form the condensations in a very short time of probably a few million years. We are working now day and night and hope to have things spick and span before I leave. …Fermi and Turkevich have recently studied in detail the formation process of the lightest elements … and found that it will not work at the densities we usually assume unless there is a strong resonance …for the reaction [⁴He + ³H] below 400 kV. Experimentally it seems that such a resonance does not exist.

Although Gamow seemed pleased that the physics of condensation of the nebulae appeared to be have been sorted out, according to Kragh “further calculations showed that the interactions were not sufficient to provide a mechanism for condensation, and Gamow was forced to give up the idea” [3, 122, 415]. And thus it was that the Gamow–Teller–Metropolis–Ulam (GTMU) collaboration remained unpublished.

For a few months in late 1948 and early 1949, Gamow worked on a short review of his recent contributions to relativistic cosmology [46]. Gamow gives scant reference to the Alpher–Herman agenda: it has one mention of a “private communication” and a cursory acknowledgment of the capture theory developed by “Alpher, Bethe, Gamow and Delter”. These mentions are examples of Gamow’s patronising manner: Alpher felt his contributions were being downplayed, while Herman was angered by the parodying of his family name. Kragh suggests, “Gamow undoubtedly enjoyed being able to smuggle a fake reference into a serious volume celebrating Einstein’s seventieth birthday” [3, 414]. Gamow scribbled a note of caution about the GTMU results to Alpher in the margin of his letter of 29 January 1949:

In that connection it may be a good idea if you take out from the m.s. on [the] expan[ding] univ[verse] the last section dealing with the formation of galaxies on my old principles, which now seem to be wrong.

Alpher and Herman did not remove “the last section” from their 1949 paper [47].

6.2 Los Alamos: 6 June 1949

At the end of the spring semester, Gamow returned to New Mexico, where Turkevich outlined his latest thinking on bridging the mass gap. Eugene Wigner had proposed that a catalytic or cyclic reaction might be the key. Wigner never published his scheme (which perhaps echoed Bethe’s CNO cycle of fusion energy in stars), so we only have Gamow to guide us: carbon-10 could catalyse lithium-6 and beryllium-7 to capture neutrons and become carbon-10 nuclei. Gamow conveyed this news in an infantile letter:

Dear Children.

Tonny Turkevitch found the way to build the “chain bridge” across the “mass-five crevasse” and is calculating now if sufficient traffic can go through. It is

₆C¹⁰ + ₁T³ → ₃Li⁶ + ₄Be⁷ + 0.002 m.u.

I hope it works!

The work on radiation pressure and galaxy-formation is now quite in order (after a few more bumps) and is going to press soon.

6.3 Los Alamos: 16 June 1949

Gamow now confessed that the GMTU paper “exists only in fragmentary parts” but he hoped to complete it within two months. That never happened. Alpher’s priority continued to be the intractable mass gap problem. Gamow suggested researchers whom Alpher should contact, namely Eugene Wigner (1902–1995) at Princeton for advice on neutron decay; Turkevich in Chicago for the latest on the mass gap; Ugo Fano (1912–2001) in Washington DC who was a great expert on the interaction of matter with radiation; and Ray Davis Jr (1912–2006) at Brookhaven for his expertise in neutrino physics. All of them were helpful suggestions.

The fourth paragraph of Gamow’s 16 June letter refers to a letter of Alpher’s which is lost. Two papers authored by the Dutch-British theoretical physicist Dirk ter Haar (1919–2002) on the origin of the chemical elements had provoked an angry reaction from the Gamow–Alpher–Herman camp [47]. The barb in Gamow’s letter to Alpher is: ter Haar is a ‘Skunck [who has gone] “back to Holland to work on a Plastic Factory. Serves him good.”¹⁰ Gamow was irritated by ter Haar’s blunt rejection of their papers on primordial nucleosynthesis: “there are many objections which can be raised against this [αβγ] theory” (several of which Alpher had already conceded) and ter Haar asserted that “the Alpher–Gamow theory would be irrelevant to the present distribution [of the elements]” [48]. Gamow probably sensed that ter Haar’s forthcoming review of stellar energy and cosmogony cast doubt on the αβγ model and the GMTU scheme [49]. Within a few months though Alpher had recovered his poise: in a thorough review of papers on primordial nucleosynthesis, Alpher and Herman (1950) cited five papers by ter Haar [50]. Gamow signed off his letter by urging Alpher: “[To] organise in APL some conference on guided missiles” in the week of 12–16 September. Such a meeting would facilitate Gamow’s travels “on Los Alamos expenses”. And as a sweetener, he added: “then we can also talk about galaxies,” the topic which remained Gamow’s number one priority.

6.4 Los Alamos: 14 November 1949

Two years after the publication of αβγ, pleasing progress had been made in modelling element formation in the ylem. Gamow had lost no time in 1948 in posting Einstein the manuscript of his short paper in Nature [43]. Einstein replied:

I am convinced that the abundance of elements as function of the atomic weight is a highly important starting point for cosmogonic speculations. The idea that the whole expansion process started with a neutron gas seems to be quite natural too. The explanation of the abundance curve by formation of the heavier elements in making use of the known facts of probability coefficients seem to me pretty convincing.¹¹

Alpher and Herman well understood by late 1949 that sceptics and rivals in other groups were investigating nucleosynthesis in the early universe. By October 1949, Alpher accepted that Fermi and Turkevich had uncovered the major stumbling block for element building by neutron capture: the mass-gap problem. Alpher gave a colloquium at which Fermi was in the audience. Fermi learned from Alpher that a lack of nuclear reaction data for the light elements was causing problems. On their return to Chicago Fermi and Turkevich generously sent their list of 28 nuclear reactions among light elements to Alpher and Herman for inclusion in their review on the origin of the elements [51, 54]. In a different approach, Maria Goeppert Mayer and Teller in Chicago suggested that although the light elements could have been formed in twenty minutes in the rapidly expanding universe, the formation of the heavy elements must have been very different. They introduced this hypothetical state of matter, a cold neutron liquid of huge size but smaller than a star. Polyneutrons, as they were dubbed, were “certainly unstable with respect to β-decay” so they would break up into heavy fragments as the population of protons and electrons in the neutron fluid increased, triggering nuclear reactions to assemble the heavy elements would be assembled [52]. They claimed the approach was “consistent with our present knowledge of isotopic abundances” [53].

George Gamow wrote to “Dear Ralph” reporting what Teller had said the month before at the Eighth Solvay conference:

I described the T&M theory in my book, and in his paper he was more conservative (drawbackish). He spoke about original continuous nuclear matter breaking up by expansion—but this is not in the paper. … It rather follows that all fragmented nuclei must be about the same size (heavy), so the distribution … would fall off on both sides … [a] Gaussian [distribution].

Alpher felt it was then impossible to discern which of the several theories would eventually prevail. Gamow signed off his letter by conceding that further progress on the condensation of galaxies had stalled, adding “it looks more and more that radiation pressure is not enough to make the trick”.

7 The team’s interests after 1949

7.1 Gamow’s interest in neutron-capture nucleosynthesis declines

On 21 November 1949, Gamow reported, “Just came back from V2 firing Carlsbad Caverns and visit to Mexico (Juarez …). Rum 75¢ a bottle!” He expected to show up in Washington on 6 February 1950 for the second semester. Gamow suggested arranging a small meeting at APL on the origin of the elements: expected his former colleague Oskar Klein to be in Washington and added that it would be nice if Klein were present. In the late 1940s, Klein was studying the physics of fluids in gravitational fields, an area that then overlapped with Gamow’s continuing interest in condensations and the origin of galaxies [54]. Gamow’s suggestion of a small workshop on the origin of elements is significant for our inquiry: Klein was openly opposed to cosmologies that posited the sudden creation of the entire universe at a given, knowable, time. Although the initial reception of αβγ had been favourable, Klein felt its shortcomings were insurmountable [55]. Gamow too now regarded further research on αβγ low priority: perhaps he was hoping Alpher would listen to Klein’s cosmological arguments and move on from αβγ.

Throughout 1950, Alpher and Herman busied themselves with further refinements to αβγ, improving the neutron-capture theory by taking into account the expansion of the universe. Alpher reported his progress at two meetings of the American Physical Society. Inclusion of the universal expansion had required “an increase by a factor of five of the density of matter chosen for the start of the element building process” [56,57,58].

7.2 Los Alamos and Pasadena: 27 May and 12 June 1951

After a gap of 18 months in the correspondence, Gamow informed his “Co-Creators” (27 May 1951) that Metropolis would provide the latest tabulation of nuclear masses in two weeks. Metropolis directed the Los Alamos team building: maniac (mathematical and numerical integrator and computer). Gamow was keen to get his collaborators to run integrations for him on. His pre-war interest in the evolution of red giant stars had revived and this letter summarises his latest thoughts on shell-burning:

The new idea is that [a shell burning red giant] will first pulsate (small and frequent central explosions which excite Cepheid pulsations), then explode more violently and less often (U Geminorum), then à la recurrent novae, ordinary nova, and finally blasts out as a supernova. This seems to fit statistically, and I am writing to [Cecilia] P. Gaposchkin for more data. In order to approach that problem, one must first have better calculation of [the] shell model, i.e. not fitting the existing fragmentary solutions by hand, but really run it through on a machine for all the percentages of mass in the core. Do you think you would be interested in doing it either on APL or on a B. of S. machine? It is probably too simple for maniac for [which] we are now coding the collapse problem.

This letter is a neat example of Gamow’s work style: it is an invitation for hands-on help with his latest foray into nuclear astrophysics. He suggested the coding of stellar collapse could be run on the IBM programmable card punch calculator or the SEAC digital computer to which Alpher had access. Alpher and Herman were far too busy with their research on neutron-capture synthesis to take on a new assignment in stellar evolution on behalf of Gamow. They had already written up their research on neutron-capture that Ralph had previewed at physics meetings [58]. Ironically, Gamow was unaware that paper had been submitted, because five days later (1 June) his second letter (from Pasadena) urged Alpher “to get it off and to go on with further developments”. Pointedly, but too late, Gamow directing his former student:

Mention the radiation density, but do not go much into details, since it certainly deserves more elaborate considerations and will take much time to clear out with all the possible consequences.

Clearly Gamow doubted “more elaborate considerations” by Alpher and Herman would lead to significant progress in the short term. The letter mentions recent research on neutron-capture theory at Caltech that Gamow felt could be relevant to the formation of condensations:

I have talked today about elementary cooking in [my] Caltech seminar and also discussed it with Mt. Wilson’s astronomers. It seems that introduction of high radiation densities may have many interesting cosmogonical consequences. Thus, for example, when the radiation was still rather dense it could help to form galactic condensations or individual stars by Spitzer-Hu[bb]le “dust-to-dust” process. … Cristy in Caltech intends to look more into the theory of capture cross-sections at mistic numbers. …

In his third letter (22 June 1951) written on his return to Los Alamos, Gamow seeks their cooperation for the modelling of the evolution of red giants.

We are all set here to start the calculations on something which will either turn out to be a Cepheid or a U Geminorum star. Do not know which! To start the calculations, we must have a composite model with contractive core and radiative envelope. Will you please rush to me the integrations on contractive core? Do you by any chance have Strömgren’s or anybody’s else solutions of radiative envelope with mass excess and mass deficiency in the center. If not we’ll have to run it here ourselves.

Alpher and Herman were unmoved this supplication.

7.3 Los Alamos 22 June 1952

A year later, Gamow suggested a collaboration with Robert Richtmyer (1910–2003):

I (and Richtmyer) are interested if you will be interested … in doing a little work which, being a preliminary to the supernova calculations, is also quite nice per se. … I am including Bob’s (Richtmyer’s) proposal on how to do it. It seems to be just a nice job for the electronic computer we have in APL. … Please look it through and let me know if you would like to do it; it would be a separate paper by Richtmyer and yourself.

PS Please return the included letter when you have copied the formulae.

After the war Teller had recruited Richtmyer, perhaps at Gamow’s bidding, to lead the Los Alamos theory division for the H-bomb project.¹² Theorists sweated round the clock on fusion calculations using ENIAC and its successor MANIAC and there was no capacity for astrophysical modelling. Alpher diligently continued with neutron-capture synthesis, showing no interest in the late stages of stellar evolution.

7.4 Santa Monica 2 June 1953

In 1950, Japanese astrophysicist Chushiro Hayashi (1920–2010) advanced the debate by including neutrinos in the hot big bang [59, 60]. Hayashi’s modelling included “beta-processes caused by energetic electrons, positrons, neutrinos and antineutrinos”. Unfortunately for Alpher and Herman the resulting neutron–proton concentration did not lead to successful generation of the relative abundances. In response they, together with James. W. Follin (1918–2011), reworked Hayashi’s calculations. In doing so, they used the latest value of the neutron decay constant (12.8 ± 2.5 min) [61] and relativistic quantum statistics. Although Gamow had no active role in this exercise, he was keen to share the results with colleagues during his summer at Los Alamos.

On 2 June 1953, he wrote from Santa Monica about his forthcoming summer visit to Los Alamos. He urgently needed:

• 1. A reprint of your article on A. Holmes;

• 2. Copy of your’s and Jim’s [Follin] article on the first five minutes

• 3. A couple of slides on each topic by your considerate choice. If you have some slides made for yourself—send them and I will return them after Ann Arbor.

The first request concerned a paper in which Alpher and Herman had estimated the primeval lead isotopic abundances and noted the agreement of such abundances with neutron capture theory. It was a beautiful example of using geochemical data to infer the evolution of radiogenic lead isotopes over cosmic time [62]. Alpher did not act on Gamow’s request for the Alpher–Follin–Herman paper because it was unfinished at that stage. However, slides of three figures from the draft, were sent.

7.5 Los Alamos: 20 June 1953

Three weeks later, Gamow thanked Alpher for the slides but confessed he did not understand them:

Thanks for the slides. They arrived safely. But I wish I would know what is plotted against what on most of them! I will use geological age slides in my first lecture. … The first five minutes comes in only on Monday July 6 which gives you time. If you cannot have your m.s. ready by that time will you send me a short letter giving the basic formulae and explanation of the slides. I will not be able to spend more than 15 or 20 min. on the subject anyway.

What are we to conclude from Gamow’s admission of a lack of understanding? Alpher had plotted many curves of density (of photons or proton-neutron ratio) versus time or temperature. Gamow seems to have lost his characteristic self-confidence. Or maybe there was a more prosaic reason for Gamow’s difficulty. From the letters (and elsewhere), one can see that after 1949 Gamow had become less interested in neutron-capture modelling. It’s possible he had not studied the neutrino physics and relativistic quantum mechanics to the level necessary for a colloquium talk, but that is mere speculation. Alpher and Herman, by contrast made good progress with theory from 1949 to 1953 while undertaking military work in their day jobs at APL/JHU.

7.6 Woods Hole, 12 June and 14 June 1955

Gamow wrote from the Marine Biological Laboratory (MBL), Woods Hole Massachusetts, two years later to say that he was doing “amusing research” on RNA, but “mostly working on my new book Matter, Earth and Sky for Prentice-Hall; the book is going great guns”. He asks Ralph for two photographs showing the pressure patterns for aircraft in and flight and for help with a diagram showing the change of drag as a function of velocity. Two days later his second letter pleads “Please do not disregard my previous letter.” Out of loyalty Alpher would surely have responded to this request, but no image is credited to him in the published book.

What was it that amused Gamow at MBL? He always delighted in puzzles posed by any new area of intellectual enquiry in which few people were working. In February 1953, Francis Crick and James Watson proposed that the molecular structure of DNA is a twisted-ladder double helix made up of long sequences of nucleotides [63]. Gamow speculated that the sequences code for the 20 amino acids that are the building blocks of proteins, quite probably being the first to propose this ingenious idea in a public forum.

7.7 New frontiers for the trio, 14 June 1955

The next letter from Woods Hole dated 14 June was the watershed for the partnership. For the three cosmogonists, it marked a parting of the ways: the lucrative Navy defense contracts had concluded. Earlier in 1955, Alpher and Herman had considered moving to the University of Iowa where their former colleague at APL/JHU, James van Allen (1914–2006) was head of the physics department. However, the low academic salaries then on offer were unattractive to young men raising families. Gamow applied to work on ballistic missiles at Convair (where Charles Critchfield was a research director), but his personal approach did not result in a job either: “I presume there must be a lot of behindthecloseddoor politics.” Gamow’s letter continued hopefully: he would soon meet with the director of the nuclear division of Glen Martin in Baltimore. That encounter did not result in a job offer for Gamow, but he found a position that was open for Alpher to consider. Alpher declined that opportunity, signing up instead with General Electric in upstate New York to work on the physics of space vehicle re-entry. Herman moved to Detroit in 1956 to join the General Motors Research Laboratory as the head of their basic science group, where he applied physics to the nascent science of traffic flow. Gamow went west to the University of Colorado Boulder where he concentrated on growing his income from highly successful science books for the general public. In his final years at Boulder, Gamow authored a dozen books, variously for Viking Press, Prentice-Hall, Doubleday, and Cambridge University Press.

After the early 1950s, professional interest in the αβγ paper faded away. The Genesis of the Big Bang autobiographical account adds on page 84:

… we published very little on nucleosynthesis calculations after 1953; we did further work with Follin, but the physical separation of Alpher, Follin, Gamow, and Herman beginning in 1955 precluded any detailed write-up of our results.

The intellectual space in which cosmologists pondered the origin of the heavy elements had time-shifted from the early universe to the late stages of the evolution of massive stars [64]. Following discovery of the CMB a revival of interest among the trio led to two further cosmological papers: in 1967 and 1968 [65, 66].

8 Conclusion

In this inquiry, we set the goal of getting a fuller understanding of how the collaborations between Gamow and, Alpher and Herman, worked in practice. Commentators other than historians have frequently praised Gamow in a lofty style of celebratory didactic narrative, of a kind derided by professionals in which glorious progress is made by heroic scientists [67, 68]. Such accounts frequently skim over the vital contributions of a supporting cast of junior colleagues, which sociologists might attribute to the Matthew effect. We note also that from the 1920s research in theoretical and observational cosmology became characterised by passionate competition within the academy. This trait persisted for two generations until precision (or concordance) cosmology flourished in the 1990s, ushering in a new spirit of international collaboration. In this review, we perceive two talented physicists with radically different personalities, one an extremely extroverted character endlessly seeking new stimulation, the other an introverted scholar who delighted in working with multivariate mathematical problems.

Through the letters, we glimpse how the collaboration worked on an everyday level. The Gamow–Alpher correspondence has enabled us to extend the Gamow’s “world line” from early 1940s to the mid-1950s, thus adding sociological detail to our understanding of his personality. Gamow sought simple answers that painted the big picture. Alpher, in the company of Herman, enjoyed the challenge of the analysis, spending weeks or months handling the operational minutiae of functions, limits, parameterisation and calculus, filling in the detail. Gamow’s letters, written in a direct and jaunty style, always address Alpher and Herman in a warm and friendly tone: what shines through in the small collection of letters is a total absence of personal rivalry and competition within the trio. Gamow’s personality and the way he interacted with Alpher is revealed in an interesting light through this collection of letters. Gamow’s career success was founded on his undoubted genius at posing seemingly impossible questions. He mostly relied on the intensive and devoted assistance of acolytes and admirers to investigate his audacious cosmological suggestions. The support of collaborators also extended to writing entire drafts of Gamow’s research papers and monographs, a commonplace arrangement between master and pupil in the academy of course. In Gamow’s case, the hyperactive master was given to understating the huge support he had received from loyal followers. Unfortunately, in the twentieth century the singular achievements of Alpher, whose contributions to the development of Gamow’s cosmology exceeded those of Gamow’s other colleagues, were largely overlooked in the literature. Fortunately, in the present century, there have been several articles that we have cited, correcting the omissions and misattributions of earlier papers, although the popular literature remains slow to respond. Our descriptive study of the Gamow–Alpher correspondence illustrates what a very remarkable partnership this was, invariably infused with friendship and the delight of scientific discovery.