History of Solar System formation and evolution hypotheses

However plausible it may appear at first sight, the nebular hypothesis still faces the obstacle of angular momentum; if the Sun had indeed formed from the collapse of such a cloud, the planets should be rotating far more slowly. The Sun, though it contains almost 99.9 percent of the system's mass, contains just 1 percent of its angular momentum.[8] This means that the Sun should be spinning much more rapidly.

Attempts to resolve the angular momentum problem led to the temporary abandonment of the nebular hypothesis in favour of a return to "two-body" theories.[7] For several decades, many astronomers preferred the tidal or near-collision hypothesis put forward by James Jeans in 1917, in which the planets were considered to have been formed due to the approach of some other star to the Sun. This near-miss would have drawn large amounts of matter out of the Sun and the other star by their mutual tidal forces, which could have then condensed into planets.[7] However, in 1929 astronomer Harold Jeffreys countered that such a near-collision was massively unlikely.[7] Objections to the hypothesis were also raised by the American astronomer Henry Norris Russell, who showed that it ran into problems with angular momentum for the outer planets, with the planets struggling to avoid being reabsorbed by the Sun.[9]

Forest Moulton in 1900 had also shown that the nebular hypothesis was inconsistent with observations because of the angular momentum. Moulton and Chamberlin in 1904 originated the planetesimal hypothesis[10] (see Chamberlin–Moulton planetesimal hypothesis). Along with many astronomers of the day they came to believe the pictures of "spiral nebulas" from the Lick Observatory were direct evidence of forming planetary systems. These turned out to be galaxies instead but the Shapley-Curtis debate about these was still 16 years in the future. One of the most fundamental issues in the history of astronomy was distinguishing between nebulas and galaxies.

Moulton and Chamberlin suggested that a star had passed close to the Sun early in its life to cause tidal bulges and that this, along with the internal process that leads to solar prominences, resulted in the ejection of filaments of matter from both stars. While most of the material would have fallen back, part of it would remain in orbit. The filaments cooled into numerous, tiny, solid fragments, ‘planetesimals’, and a few larger protoplanets. This model received favourable support for about 3 decades but passed out of favour by the late '30s and was discarded in the '40s by the realization it was incompatible with the angular momentum of Jupiter, but a part of it, planetesimal accretion, was retained.[3]

In 1937 and 1940, Ray Lyttleton postulated that a companion star to the Sun collided with a passing star.[3] Such a scenario was already suggested and rejected by Henry Russell in 1935 (though it may have been more likely assuming the Sun was born in an open cluster, where stellar collisions are common). Lyttleton showed terrestrial planets were too small to condense on their own so suggested one very large proto-planet broke in two because of rotational instability, forming Jupiter and Saturn, with a connecting filament from which the other planets formed. A later model, from 1940 and 1941, involves a triple star system, a binary plus the Sun, in which the binary merges and later breaks up because of rotational instability and escapes from the system leaving a filament that formed between them to be captured by the Sun. Objections of Lyman Spitzer apply to this model also.

In 1954, 1975, and 1978[11] Swedish astrophysicist Hannes Alfvén included electromagnetic effects in equations of particle motions, and angular momentum distribution and compositional differences were explained. In 1954 he first proposed the band structure in which he distinguished an A-cloud, containing mostly helium, but with some solid- particle impurities ("meteor rain"), a B-cloud, with mostly carbon, a C-cloud, having mainly hydrogen, and a D-cloud, made mainly of silicon and iron. Impurities in the A-cloud form Mars and the Moon (later captured by Earth), impurities in the B-cloud collapse to form the outer planets, in the C-cloud they condense into Mercury, Venus, Earth, the asteroid belt, moons of Jupiter, and Saturn's rings, while Pluto, Triton, the outer satellites of Saturn, the moons of Uranus, the Kuiper Belt, and Oort cloud may have formed from the D-cloud.

In 1943, the Soviet astronomer Otto Schmidt proposed that the Sun, in its present form, passed through a dense interstellar cloud, emerging enveloped in a cloud of dust and gas, from which the planets eventually formed. This solved the angular momentum problem by assuming that the Sun's slow rotation was peculiar to it, and that the planets did not form at the same time as the Sun.[7] Extensions of the model, together forming the Russian school, include Gurevich and Lebedinsky (in 1950), Safronov (in 1967,1969), Safronov and Vityazeff (in 1985), Safronov and Ruskol (in 1994), and Ruskol (in 1981), among others[12] However, this hypothesis was severely dented by Victor Safronov who showed that the amount of time required to form the planets from such a diffuse envelope would far exceed the Solar System's determined age.[7]

Ray Lyttleton modified the theory by showing that a 3rd body was not necessary and proposing that a mechanism of line accretion described by Bondi and Hoyle in 1944 would enable cloud material to be captured by the star (Williams and Cremin, 1968, loc. cit.)

In this model[3] (from 1944) the companion went nova with ejected material captured by the Sun and planets forming from this material. In a version a year later it was a supernova. In 1955 he proposed a similar system to Laplace, and with more mathematical detail in 1960. It differs from Laplace in that a magnetic torque occurs between the disk and the Sun, which comes into effect immediately or else more and more matter would be ejected resulting in a much too massive planetary system, one comparable to the Sun. The torque causes a magnetic coupling and acts to transfer angular momentum from the Sun to the disk. The magnetic field strength would have to be 1 gauss. The existence of torque depends on magnetic lines of force being frozen into the disk (a consequence of a well-known MHD (magnetohydrodynamic) theorem on frozen-in lines of force). As the solar condensation temperature when the disk was ejected could not be much more than 1000 degrees K., a number of refractories must be solid, probably as fine smoke particles, which would grow with condensation and accretion. These particles would be swept out with the disk only if their diameter at the Earth's orbit was less than 1 meter so as the disk moved outward a subsidiary disk consisting of only refractories remains behind where the terrestrial planets would form. The model is in good agreement with the mass and composition of the planets and angular momentum distribution provided the magnetic coupling is an acceptable idea, but not explained are twinning, the low mass of Mars and Mercury, and the planetoid belts. It was Alfvén who formulated the concept of frozen-in magnetic field lines.

Gerard Kuiper (in 1944)[3] argued, like Ter Haar, that regular eddies would be impossible and postulated that large gravitational instabilities might occur in the solar nebula, forming condensations. In this, the solar nebula could be either co-genetic with the Sun or captured by it. Density distribution would determine what could form: either a planetary system or a stellar companion. The 2 types of planets were assumed to be due to the Roche limit. No explanation was offered for the Sun's slow rotation which Kuiper saw as a larger G-star problem.

In Fred Whipple's 1948 scenario[3] a smoke cloud about 60,000 AU in diameter and with 1 solar mass (M_☉) contracts and produces the Sun. It has a negligible angular momentum thus accounting for the Sun's similar property. This smoke cloud captures a smaller one with a large angular momentum. The collapse time for the large smoke and gas nebula is about 100 million years and the rate is slow at first, increasing in later stages. The planets would condense from small clouds developed in, or captured by, the 2nd cloud, the orbits would be nearly circular because accretion would reduce eccentricity due to the influence of the resisting medium, orbital orientations would be similar because the small cloud was originally small and the motions would be in a common direction. The protoplanets might have heated up to such high degrees that the more volatile compounds would have been lost and the orbital velocity decreases with increasing distance so that the terrestrial planets would have been more affected. The weaknesses of this scenario are that practically all the final regularities are introduced as a prior assumptions and most of the hypothesizing was not supported by quantitative calculations. For these reasons it did not gain wide acceptance.

American chemist Harold Urey, who founded cosmochemistry, put forward a scenario[3] in 1951, 1952, 1956, and 1966 based largely on meteorites and using Chandrasekhar's stability equations and obtained density distribution in the gas and dust disk surrounding the primitive Sun. In order that volatile elements like mercury could be retained by the terrestrial planets he postulated a moderately thick gas and dust halo shielding the planets from the Sun. In order to form diamonds, pure carbon crystals, Moon-size objects, gas spheres that became gravitationally unstable, would have to form in the disk with the gas and dust dissipating at a later stage. Pressure fell as gas was lost and diamonds were converted to graphite, while the gas became illuminated by the Sun. Under these conditions considerable ionization would be present and the gas would be accelerated by magnetic fields, hence the angular momentum could be transferred from the Sun. He postulated that these lunar-size bodies were destroyed by collisions, with the gas dissipating, leaving behind solids collected at the core, with the resulting smaller fragments pushed far out into space and the larger fragments staying behind and accreting into planets. He suggested the Moon was just such a surviving core.

In 1960, 1963, and 1978,[13] W. H. McCrea proposed the protoplanet theory, in which the Sun and planets individually coalesced from matter within the same cloud, with the smaller planets later captured by the Sun's larger gravity.[7] It includes fission in a protoplanetary nebula and there is no solar nebula. Agglomerations of floccules (which are presumed to compose the supersonic turbulence assumed to occur in the interstellar material from which stars are born) formed the Sun and protoplanets, the latter splitting to form planets. The 2 portions can not remain gravitationally bound to each other, are at a mass ratio of at least 8 to 1, and for inner planets they go into independent orbits while for outer planets one of the portions exits the Solar System. The inner protoplanets were Venus-Mercury and Earth-Mars. The moons of the greater planets were formed from "droplets" in the neck connecting the 2 portions of the dividing protoplanet and these droplets could account for some of the asteroids. Terrestrial planets would have no major moons which does not account for Luna. It predicts certain observations such as the similar angular velocity of Mars and Earth with similar rotation periods and axial tilts. In this scheme there are 6 principal planets: 2 terrestrial, Venus and Earth, 2 major, Jupiter and Saturn, and 2 outer, Uranus and Neptune; and 3 lesser planets: Mercury, Mars, and Pluto.

This theory has a number of problems, such as explaining the fact that the planets all orbit the Sun in the same direction with relatively low eccentricity, which would appear highly unlikely if they were each individually captured.[7]

In American astronomer Alastair G. W. Cameron's hypothesis (from 1962 and 1963),[3] the protosun has a mass of about 1–2 Suns with a diameter of around 100,000 AU is gravitationally unstable, collapses, and breaks up into smaller subunits. The magnetic field is of the order of 1/100,000 gauss. During the collapse the magnetic lines of force are twisted. The collapse is fast and is done by the dissociation of H molecules followed by the ionization of H and the double ionization of He. Angular momentum leads to rotational instability which produces a Laplacean disk. At this stage radiation will remove excess energy and the disk will be quite cool in a relatively short period (about 1 mln. yrs.) and the condensation into what Whipple calls cometismals takes place. Aggregation of these produces giant planets which in turn produce disks during their formation from which evolve into lunar systems. The formation of terrestrial planets, comets, and asteroids involved disintegration, heating, melting, solidification, etc. He also formulated the giant-impact hypothesis for the origin of the Moon.

The capture theory, proposed by Michael Mark Woolfson in 1964, posits that the Solar System formed from tidal interactions between the Sun and a low-density protostar. The Sun's gravity would have drawn material from the diffuse atmosphere of the protostar, which would then have collapsed to form the planets.[14] However, the capture theory predicts a different age for the Sun than for the planets, whereas the similar ages of the Sun and the rest of the Solar System indicate that they formed at roughly the same time.[15]

As captured planets would have initially eccentric orbits Dormand and Woolfson in 1974 and 1977 and Woolfson[16] proposed the possibility of a collision. A filament is thrown out by a passing proto-star which is captured by the Sun and planets form from it. In this there were 6 original planets, corresponding to 6 point-masses in the filament, with planets "Enyo" and "Bellona", the 2 innermost, colliding. Enyo, at twice the mass of Neptune, is ejected out of the Solar System, while Bellona, estimated to be a third the mass of Uranus, splits into two to form Earth and Venus. In a revised (2017) version of the hypothesis, Bellona and Enyo are now both gas giants more massive than Jupiter and their collision briefly causes deuterium-deuterium chain reactions, shattering both planets. Sediments from Enyo's interior form Venus while sediments from Bellona's interior form Earth.[17] Mars, the Moon, Haumea, Makemake, Eris, and V774104 are former moons of Enyo. Mercury is either a fragment of Bellona or an escaped moon of Enyo. The Enyo-Bellona collision also formed the asteroid belt, Kuiper belt, Oort cloud, and comets. Pluto, either a fragment or moon of one of the planets, passed close to Neptune's satellite Triton causing it to assume its retrograde orbit.[18]

T.J.J. See was an American astronomer and Navy Captain who at one time worked under Ellery Hale at the Lowell Observatory. He had a cult following largely because of his many (some 60) articles in Popular Astronomy but also in Astronomische Nachrichte (Astronomical News) (mostly in English). While at the USNO's Mare Island, Cal. station, he developed a model which he called capture theory, published in 1910, in his "Researches on the Evolution of the Stellar Systems: v. 2. The capture theory of cosmical evolution, founded on dynamical principles and illustrated by phenomena observed in the spiral nebulae, the planetary system, the double and multiple stars and clusters and the star-clouds of the Milky Way", which proposed that the planets formed in the outer Solar System and were captured by the Sun; the moons were formed in thus manner and were captured by the planets. This caused a feud with Forest Moulton, who co-developed the planetesimal hypothesis. A preview was presented in 1909 at a meeting of the ASP (Astronomical Society of the Pacific) at the Chabot Observatory in Oakland, Cal., and newspaper headlines blared "Prof. See's Paper Causes Sensation" (San Francisco Call) and "Scientists in Furore Over Nebulae" (San Francisco Examiner). Our current knowledge of dynamics makes capture most unlikely as it requires special conditions.[10]

Swiss astronomer Louis Jacot (in 1951, 1962, 1981),[19] like Weisacker and Ter Haar, continued the Cartesian idea of vortices but proposed a hierarchy of vortices or vortices within vortices, i.e., a lunar system vortex, a Solar System vortex, and a galactic vortex. He put forward the notion that planetary orbits are spirals, not circles or ellipses. Jacot also proposed the expansion of galaxies (stars move away from the hub), and that moons move away from their planets.

He also maintained that planets were expelled, one at a time, from the Sun, specifically from an equatorial bulge caused by rotation, and that one of them shattered in this expulsion leaving the asteroid belt. The Kuiper Belt was unknown at the time, but presumably it, too, would be the result of the same kind of shattering. The moons, like the planets, originated as equatorial expulsions, but, of course, from their parent planets, with some shattering, leaving the rings, and Earth is supposed to eventually expel another moon.

In this model there were 4 phases to the planets: no rotation and keeping the same side to the Sun "as Mercury does now" (we've known, of course, since 1965, that it doesn't), very slow, accelerated, and finally, daily rotation.

He explained the differences between inner and outer planets and inner and outer moons through vortex behaviour. Mercury's eccentric orbit was explained by its recent expulsion from the Sun and Venus' slow rotation as its being in the "slow rotation phase", having been expelled second to last.

The Tom Van Flandern model[20][21][22][23] was first proposed in 1993 in the first edition of his book. In the revised version from 1999 and later, the original Solar System had 6 pairs of twin planets each fissioned off from the equatorial bulges of an overspinning Sun (outward centrifugal forces exceed the inward gravitational force) at different times so having different temperatures, sizes, and compositions, and having condensed thereafter with the nebular disk dissipating after some 100 million years, with 6 planets exploding. Four of these were helium dominated, fluid, and unstable (helium class planets). These were V (Maldek)[24] (V standing for the 5th planet, the first 4 including Mercury and Mars), K (Krypton), T (transneptunian), and Planet X. In these cases, the smaller moons exploded because of tidal stresses leaving the 4 component belts of the 2 major planetoid zones. Planet LHB-A, the explosion for which is postulated to have caused the Late Heavy Bombardment (about 4 eons ago), was twinned with Jupiter, and LHB-B, the explosion for which is postulated to have caused another LHB, was twinned with Saturn. In planets LHB-A, Jupiter, LHB-B, and Saturn, being gigantic, Jovian planets, the inner and smaller partner in each pair was subjected to enormous tidal stresses causing it to blow up. The explosions took place before they were able to fission off moons. As the 6 were fluid they left no trace. Solid planets fission off only one moon and Mercury was a moon of Venus but drifted away because of the Sun's gravitational influence. Mars was a moon of Maldek.

One major argument against exploding planets and moons is that there would not be an energy source powerful enough to cause such explosions.

In J. Marvin Herndon's model,[25] inner (large-core) planets form by condensation and raining-out from within giant gaseous protoplanets at high pressures and high temperatures. Earth's complete condensation included a c. 300 Earth-mass gas/ice shell that compressed the rocky kernel to about 66% of Earth's present diameter (Jupiter equates to about 300 Earth masses, which equals c. 2000 trillion trillion kg; Earth is at about 6 trillion trillion kg). T Tauri (see T Tauri type stars) eruptions of the Sun stripped the gases away from the inner planets. Mercury was incompletely condensed and a portion of its gases were stripped away and transported to the region between Mars and Jupiter, where it fused with in-falling oxidized condensate from the outer reaches of the Solar System and formed the parent material for ordinary chondrite meteorites, the Main-Belt asteroids, and veneer for the inner planets, especially Mars. The differences between the inner planets are primarily the consequence of different degrees of protoplanetary compression. There are two types of responses to decompression-driven planetary volume increases: cracks, which form to increase surface area, and folding, creating mountain ranges, to accommodate changes in curvature.

This planetary formation theory represents an extension of the Whole-Earth Decompression Dynamics (WEDD) model,[26] which includes natural nuclear-fission reactors in planetary cores; Herndon elaborates, expounds, and elucidates it in 11 articles in Current Science from 2005 to 2013 and in five books published from 2008 to 2012. He refers to his model as "indivisible" – meaning that the fundamental aspects of Earth are connected logically and causally, and can be deduced from its early formation as a Jupiter-like giant.

In 1944 the German chemist and physicist Arnold Eucken considered the thermodynamics of Earth condensing and raining-out within a giant protoplanet at pressures of 100–1000 atm. In the 1950s and early 1960s discussion of planetary formation at such pressures took place, but Cameron's 1963 low-pressure (c. 4–10 atm.) model largely supplanted the idea.

Jeans, in 1931, divided the various models into 2 groups: those where the material for planet formation came from the Sun and those where it didn't and may be concurrent or consecutive.[27]

William McCrea, in 1963, divided them into 2 groups also: those that relate the formation of the planets to the formation of the Sun and those where it is independent of the formation of the Sun, where the planets form after the Sun becomes a normal star.[27]

Ter Haar and Cameron[28] distinguished between those theories that consider a closed system, which is a development of the Sun and possibly a solar envelope, that starts with a protosun rather than the Sun itself, and state that Belot calls these theories monistic; and those that consider an open system, which is where there is an interaction between the Sun and some foreign body that is supposed to have been the first step in the developments leading to the planetary system, and state that Belot calls these theories dualistic.

Hervé Reeves' classification[29] also categorizes them as co-genetic with the Sun or not but also as formed from altered or unaltered stellar/interstellar material. He as well recognizes 4 groups: 1) models based on the solar nebula, originated by Swedenborg, Kant, and Laplace in the 1700s; 2) the ones proposing a cloud captured from interstellar space, major proponents being Alfvén and Gustaf Arrhenius (in 1978) and Alfvén and Arrhenius; 3) the binary hypotheses which propose that a sister star somehow disintegrated and a portion of its dissipating material was captured by the Sun, principal hypothesizer being Lyttleton in the '40s; 4) and the close-approach-filament ideas of Jeans, Jeffreys, and Woolfson and Dormand.

In Williams and Cremin[27] the categories are: (1) models that regard the origin and formation of the planets as being essentially related to the Sun, with the 2 formation processes taking place concurrently or consecutively, (2) models that regard formation of the planets as being independent of the formation process of the Sun, the planets forming after the Sun becomes a normal star; this has 2 subcategories: a) where the material for the formation of the planets is extracted either from the Sun or another star, b) where the material is acquired from interstellar space. They conclude that the best models are Hoyle's magnetic coupling and McCrea's floccules.

Woolfson[30] recognized 1) monistic, which included Laplace, Descartes, Kant, and Weisacker, and 2) dualistc, which included Leclerc (comte de Buffon), Chamberlin-Moulton, Jeans, Jeffreys, and Schmidt-Lyttleton.

Beta Pictoris seen by the Hubble Space Telescope

In 1978, astronomer A. J. R. Prentice revived the Laplacian nebular model in his Modern Laplacian Theory by suggesting that the angular momentum problem could be resolved by drag created by dust grains in the original disc which slowed down the rotation in the centre.[7][31] Prentice also suggested that the young Sun transferred some angular momentum to the protoplanetary disc and planetesimals through supersonic ejections understood to occur in T Tauri stars.[7][32] However, his contention that such formation would occur in toruses or rings has been questioned, as any such rings would disperse before collapsing into planets.[7]

The birth of the modern widely accepted theory of planetary formation—the Solar Nebular Disk Model (SNDM)—can be traced to the works of Soviet astronomer Victor Safronov.[33] His book Evolution of the protoplanetary cloud and formation of the Earth and the planets,[34] which was translated to English in 1972, had a long-lasting effect on the way scientists thought about the formation of the planets.[35] In this book almost all major problems of the planetary formation process were formulated and some of them solved. Safronov's ideas were further developed in the works of George Wetherill, who discovered runaway accretion.[7] By the early 1980s, the nebular hypothesis in the form of SNDM had come back into favour, led by two major discoveries in astronomy. First, a number of apparently young stars, such as Beta Pictoris, were found to be surrounded by discs of cool dust, much as was predicted by the nebular hypothesis. Second, the Infrared Astronomical Satellite, launched in 1983, observed that many stars had an excess of infrared radiation that could be explained if they were orbited by discs of cooler material.

While the broad picture of the nebular hypothesis is widely accepted,[36] many of the details are not well understood and continue to be refined.

The refined nebular model was developed entirely on the basis of observations of the Solar System because it was the only one known until the mid-1990s. It was not confidently assumed to be widely applicable to other planetary systems, although scientists were anxious to test the nebular model by finding of protoplanetary discs or even planets around other stars.[37] As of August 30, 2013, the discovery of 941 extrasolar planets[38] has turned up many surprises, and the nebular model must be revised to account for these discovered planetary systems, or new models considered.

Among the extrasolar planets discovered to date are planets the size of Jupiter or larger but possessing very short orbital periods of only a few hours. Such planets would have to orbit very closely to their stars; so closely that their atmospheres would be gradually stripped away by solar radiation.[39][40] There is no consensus on how to explain these so-called hot Jupiters, but one leading idea is that of planetary migration, similar to the process which is thought to have moved Uranus and Neptune to their current, distant orbit. Possible processes that cause the migration include orbital friction while the protoplanetary disk is still full of hydrogen and helium gas[41] and exchange of angular momentum between giant planets and the particles in the protoplanetary disc.[42][43][44]

The detailed features of the planets are another problem. The solar nebula hypothesis predicts that all planets will form exactly in the ecliptic plane. Instead, the orbits of the classical planets have various (but small) inclinations with respect to the ecliptic. Furthermore, for the gas giants it is predicted that their rotations and moon systems will also not be inclined with respect to the ecliptic plane. However, most gas giants have substantial axial tilts with respect to the ecliptic, with Uranus having a 98° tilt.[45] The Moon being relatively large with respect to the Earth and other moons which are in irregular orbits with respect to their planet is yet another issue. It is now believed these observations are explained by events which happened after the initial formation of the Solar System.[46]

At that time, the prevailing scientific view on the source of the Sun's heat was that it was generated by gravitational contraction. In the 1840s, astronomers J. R. Mayer and J. J. Waterson first proposed that the Sun's massive weight causes it to collapse in on itself, generating heat, an idea expounded upon in 1854 by both Hermann von Helmholtz and Lord Kelvin, who further elaborated on the idea by suggesting that heat may also be produced by the impact of meteors onto the Sun's surface.[47] Theories at the time suggested that stars evolved moving down the main-sequence of the Hertzsprung-Russell diagram, starting off as diffuse red supergiants before contracting and heating to become blue main-sequence stars, then even further down to red dwarfs before finally ending up as cool, dense black dwarfs. However, the Sun only has enough gravitational potential energy to power its luminosity by this mechanism for about 30 million years—far less than the age of the Earth. (This collapse time is known as the Kelvin–Helmholtz timescale.)[48]

Albert Einstein's development of the theory of relativity in 1905 led to the understanding that nuclear reactions could create new elements from smaller precursors, with the loss of energy. In his treatise Stars and Atoms, Arthur Eddington suggested that pressures and temperatures within stars were great enough for hydrogen nuclei to fuse into helium; a process which could produce the massive amounts of energy required to power the Sun.[47] In 1935, Eddington went further and suggested that other elements might also form within stars.[49] Spectral evidence collected after 1945 showed that the distribution of the commonest chemical elements, carbon, hydrogen, oxygen, nitrogen, neon, iron etc., was fairly uniform across the galaxy. This suggested that these elements had a common origin.[49] A number of anomalies in the proportions hinted at an underlying mechanism for creation. Lead has a higher atomic weight than gold, but is far more common. Hydrogen and helium (elements 1 and 2) are virtually ubiquitous yet lithium and beryllium (elements 3 and 4) are extremely rare.[49]

While the unusual spectra of red giant stars had been known since the 19th century,[50] it was George Gamow who, in the 1940s, first understood that they were stars of roughly solar mass that had run out of hydrogen in their cores and had resorted to burning the hydrogen in their outer shells. This allowed Martin Schwarzschild to draw the connection between red giants and the finite lifespans of stars. It is now understood that red giants are stars in the last stages of their life cycles.

Fred Hoyle noted that, even while the distribution of elements was fairly uniform, different stars had varying amounts of each element. To Hoyle, this indicated that they must have originated within the stars themselves. The abundance of elements peaked around the atomic number for iron, an element that could only have been formed under intense pressures and temperatures. Hoyle concluded that iron must have formed within giant stars.[49] From this, in 1945 and 1946, Hoyle constructed the final stages of a star's life cycle. As the star dies, it collapses under its own weight, leading to a stratified chain of fusion reactions: carbon-12 fuses with helium to form oxygen-16; oxygen-16 fuses with helium to produce neon-20, and so on up to iron.[51] There was, however, no known method by which carbon-12 could be produced. Isotopes of beryllium produced via fusion were too unstable to form carbon, and for three helium atoms to form carbon-12 was so unlikely as to have been impossible over the age of the Universe. However, in 1952 the physicist Ed Salpeter showed that a short enough time existed between the formation and the decay of the beryllium isotope that another helium had a small chance to form carbon, but only if their combined mass/energy amounts were equal to that of carbon-12. Hoyle, employing the anthropic principle, showed that it must be so, since he himself was made of carbon, and he existed. When the matter/energy level of carbon-12 was finally determined, it was found to be within a few percent of Hoyle's prediction.[52]

The first white dwarf discovered was in the triple star system of 40 Eridani, which contains the relatively bright main sequence star 40 Eridani A, orbited at a distance by the closer binary system of the white dwarf 40 Eridani B and the main sequence red dwarf 40 Eridani C. The pair 40 Eridani B/C was discovered by William Herschel on January 31, 1783;[53]^{, p. 73} it was again observed by Friedrich Georg Wilhelm Struve in 1825 and by Otto Wilhelm von Struve in 1851.[54][55] In 1910, it was discovered by Henry Norris Russell, Edward Charles Pickering and Williamina Fleming that despite being a dim star, 40 Eridani B was of spectral type A, or white.[56]

White dwarfs were found to be extremely dense soon after their discovery. If a star is in a binary system, as is the case for Sirius B and 40 Eridani B, it is possible to estimate its mass from observations of the binary orbit. This was done for Sirius B by 1910,[57] yielding a mass estimate of 0.94 M_☉. (A more modern estimate is 1.00 M_☉.)[58] Since hotter bodies radiate more than colder ones, a star's surface brightness can be estimated from its effective surface temperature, and hence from its spectrum. If the star's distance is known, its overall luminosity can also be estimated. Comparison of the two figures yields the star's radius. Reasoning of this sort led to the realization, puzzling to astronomers at the time, that Sirius B and 40 Eridani B must be very dense. For example, when Ernst Öpik estimated the density of a number of visual binary stars in 1916, he found that 40 Eridani B had a density of over 25,000 times the Sun's, which was so high that he called it "impossible".[59]

Such densities are possible because white dwarf material is not composed of atoms bound by chemical bonds, but rather consists of a plasma of unbound nuclei and electrons. There is therefore no obstacle to placing nuclei closer to each other than electron orbitals—the regions occupied by electrons bound to an atom—would normally allow.[60] Eddington, however, wondered what would happen when this plasma cooled and the energy which kept the atoms ionized was no longer present.[61] This paradox was resolved by R. H. Fowler in 1926 by an application of the newly devised quantum mechanics. Since electrons obey the Pauli exclusion principle, no two electrons can occupy the same state, and they must obey Fermi–Dirac statistics, also introduced in 1926 to determine the statistical distribution of particles which satisfy the Pauli exclusion principle.[62] At zero temperature, therefore, electrons could not all occupy the lowest-energy, or ground, state; some of them had to occupy higher-energy states, forming a band of lowest-available energy states, the Fermi sea. This state of the electrons, called degenerate, meant that a white dwarf could cool to zero temperature and still possess high energy.

Planetary nebulae are generally faint objects, and none are visible to the naked eye. The first planetary nebula discovered was the Dumbbell Nebula in the constellation of Vulpecula, observed by Charles Messier in 1764 and listed as M27 in his catalogue of nebulous objects. To early observers with low-resolution telescopes, M27 and subsequently discovered planetary nebulae somewhat resembled the gas giants, and William Herschel, discoverer of Uranus, eventually coined the term 'planetary nebula' for them, although, as we now know, they are very different from planets.

The central stars of planetary nebulae are very hot. Their luminosity, though, is very low, implying that they must be very small. Only once a star has exhausted all its nuclear fuel can it collapse to such a small size, and so planetary nebulae came to be understood as a final stage of stellar evolution. Spectroscopic observations show that all planetary nebulae are expanding, and so the idea arose that planetary nebulae were caused by a star's outer layers being thrown into space at the end of its life.

However, these hypotheses were all refuted by the late 1960s and early 1970s Apollo lunar missions, which introduced a stream of new scientific evidence; specifically concerning the Moon's composition, its age, and its history. These lines of evidence contradict many predictions made by these earlier models.[63] The rocks brought back from the Moon showed a marked decrease in water relative to rocks elsewhere in the Solar System, and also evidence of an ocean of magma early in its history, indicating that its formation must have produced a great deal of energy. Also, oxygen isotopes in lunar rocks showed a marked similarity to those on Earth, suggesting that they formed at a similar location in the solar nebula. The capture model fails to explain the similarity in these isotopes (if the Moon had originated in another part of the Solar System, those isotopes would have been different), while the co-accretion model cannot adequately explain the loss of water (if the Moon formed in a similar fashion to the Earth, the amount of water trapped in its mineral structure would also be roughly similar). Conversely, the fission model, while it can account for the similarity in chemical composition and the lack of iron in the Moon, cannot adequately explain its high orbital inclination and, in particular, the large amount of angular momentum in the Earth–Moon system, more than any other planet–satellite pair in the Solar System.[63]

For many years after Apollo, the binary accretion model was settled on as the best hypothesis for explaining the Moon's origins, even though it was known to be flawed. Then, at a conference in Kona, Hawaii in 1984, a compromise model was composed that accounted for all of the observed discrepancies. Originally formulated by two independent research groups in 1976, the giant impact model supposed that a massive planetary object, the size of Mars, had collided with Earth early in its history. The impact would have melted Earth's crust, and the other planet's heavy core would have sunk inward and merged with Earth's. The superheated vapour produced by the impact would have risen into orbit around the planet, coalescing into the Moon. This explained the lack of water (the vapour cloud was too hot for water to condense), the similarity in composition (since the Moon had formed from part of the Earth), the lower density (since the Moon had formed from the Earth's crust and mantle, rather than its core), and the Moon's unusual orbit (since an oblique strike would have imparted a massive amount of angular momentum to the Earth–Moon system).[63]

However, the giant impact model has been criticised for being too explanatory; it can be expanded to explain any future discoveries and as such, is unfalsifiable. Also, many claim that much of the material from the impactor would have ended up in the Moon, meaning that the isotope levels would be different, but they are not. Also, while some volatile compounds such as water are absent from the Moon's crust, many others, such as manganese, are not.[63]

While the co-accretion and capture models are not currently accepted as valid explanations for the existence of the Moon, they have been employed to explain the formation of other natural satellites in the Solar System. Jupiter's Galilean satellites are believed to have formed via co-accretion,[64] while the Solar System's irregular satellites, such as Triton, are all believed to have been captured.[65]