Asteroids are concentrated in two main regions: a wide belt between the orbits of Mars and Jupiter, and the orbit of Jupiter itself, where a substantial number of asteroids, called Trojans, have drifted into synchrony with Jupiter’s motion. Could they be the remains of former planets? Models of the early solar system suggest that the main asteroid belt contains no more than 0.5% of the material that was present initially, and there was certainly room. The vacancy extends across a radial distance more than twice that between Mars and the Sun, which accommodates four terrestrial planets. In the region between Mars, the outermost terrestrial, and Jupiter, the innermost gaseous planet, there could have been one, two, or even more.

Another vacancy lies between Uranus and Neptune. Distances in the solar system are measured in astronomical units or AU, one unit being defined as the mean distance between the Sun and the Earth. The orbits of Uranus and Neptune are 11 AU apart, twice the entire distance separating the Sun and Jupiter. With Uranus’s axis tilted 98 degrees it certainly seems as if something catastrophic occurred in this region. Was the axis knocked sideways by a massive asteroid?

All but one of the planets orbit within 3.5 degrees of a single plane, called the plane of the ecliptic. The exception is Mercury, with an inclination of 7 degrees; its orbit is also the least circular. Asteroids tend to orbit at a much more inclined angle: the majority within 20 degrees of the plane, a few up to 40 degrees. In the nebula hypothesis, a 3-dimensional cloud collapses into a planar disc and the planetary bodies accreting out of it orbit close to the plane. In the design hypothesis, their original orbits would be exactly on the same plane (since any randomness would imply, misleadingly, a physical cause for the randomness). In either case, any significant deviation from the plane implies a measure of disturbance. In the case of asteroids, the disturbance is due to collisions: repeated jostling within the belt results in cycles of break-up and re-aggregation that continue to the present day. Even on quite small asteroids impact scars are common. In the case of planets, the disturbance is due to primeval mega-impacts, collectively known as the ‘Late Heavy Bombardment’. The Moon is covered in impact craters from this event, the largest of them 2,500 km across.

Fragments from the explosions shot out in all directions. Those that shot out in roughly the same direction as the orbit of the former planet would have remained at that distance. Fragments that shot out in other directions would (i) sooner or later have crashed into the Sun or other planets, (ii) been captured by other planets to become their satellites, or (iii) have attained solar orbits of their own.

Much of the material inferred to be missing from the asteroid belt was swallowed up by Jupiter, immediately adjacent, which thereby acquired a core equal, it is estimated, to about 16 Earths. The Trojan asteroids and the rocky moons of Jupiter and Saturn, which all have similar optical colours, may likewise have originated from planetary explosions between Mars and Jupiter. At the other extreme we have the constituents of the Sun itself to consider. Initially the Sun would have consisted solely of hydrogen, which it has since been turning into helium, but most of its other elements exist in chondrite-like proportions, as a result of planetary debris having plunged into it. Taking into account all the out-of-place rocky material in the solar system, it becomes apparent that only a tiny fraction of the mass of the original planets remains in the asteroid belt.

Comets

As with every other class of solar system body, ideas about comets have changed over recent years. Once assumed to be unaltered leftovers of the cold outer nebula, they appear on closer inspection to be compositionally much the same as asteroids, only richer in volatiles (i.e. ices). We know most about the short-period variety: those with orbits of less than 200 years. Some consist primarily of volatiles; some have solid nuclei, dented by craters; others are low-density rubble piles. The bodies become comets when their elliptical orbits bring them close to the Sun; they permanently cease to be comets when they have lost their volatiles. Each year several asteroids are reclassified as comets on being found to exhibit a coma and/or tail. Most short-period comets orbit between Mars and Jupiter, having been drawn into the inner solar system, it is supposed, from the Kuiper Belt.

The data sent back by the recent Deep Impact and Stardust missions, investigating Tempel 1 and Wild 2 respectively, have been revelatory. Contrary to the expectation that the silicate particles amidst the ices would be amorphous, like the dust in interstellar space, it turned out that the silicates were crystalline, indicating subjection to temperatures up to 1,000 ºC. Analysis of one such silicate in Wild 2 indicated a cooling rate within the range 10–100 °C per hour, in close agreement with the cooling rates of chondrules or asteroidal lava (Leroux et al 2008). Carbonates, iron, CAIs and chondrules were also detected.

How is one to explain the existence of high-temperature materials and low-temperature materials in the selfsame body? Since the refractory core must have preceded the volatiles around the core, it is the core that must play the decisive role in resolving the problem, and the crystalline rocky material in the core suggests an origin in the inner solar system. Like asteroids, short-period comets seem to have originated as fragments of terrestrial planets, mixing with with water vapour as they sped out.

‘Long-period’ comets are those with elliptical orbits that take upwards of 200 years to complete, extending far beyond the observable solar system. They too seem not to be fundamentally different from asteroids in composition. Their life-times are long in comparison to short-period comets but short in comparison to the presumed age of the solar system. They can orbit the Sun at any inclination and are as likely to have a retrograde orbit as a prograde one. These characteristics have prompted the hypothesis that the solar system is enclosed by a vast replenishing ‘cloud’ of such comets, known as the ‘Oort cloud’, at a distance too great to permit the hypothesis to be corroborated. Whatever the reality of the cloud, the difference in orbital period between the two types is not absolute. Long-period comets can become short-period if the gravity of the giant planets eventually pulls them into the inner solar system – i.e. pulls them back into the inner solar system. The Centaurs, little known half-comet, half-asteroid objects between Neptune and Jupiter, may be related to the long-period comets. They are of particular interest because their orbits are strongly chaotic, i.e. they vary markedly on timescales of a few thousand years.

Moons

Except for the Earth’s moon, none of the planetary satellites are likely to have formed in situ. Mercury and Venus have no satellites. Mars has two, both clearly asteroids. Jupiter has at least 63 and Saturn at least 62, the majority less than 10 km in diameter. Uranus has 27, Neptune 13. In the scenario proposed here, satellites, depending on their composition, are (i) expelled main belt asteroids, (ii) debris from an exploded gaseous planet between the orbits of Uranus and Neptune (a possible origin of some long-period comets), or (iii) expelled Kuiper Belt Objects. Once captured, the different types frequently crashed into each other. Some collisions resulted in complete fragmentation. Occasionally the fragments may have re-assembled to form new moons. Other collisions merely left the moons cratered and scarred. Spherical shape is not evidence of a particular origin, since above a certain mass self-gravity will naturally pull a body into a sphere.

Astronomers classify the moons into regular satellites, which trace close, circular orbits about the host’s equator, and irregular satellites, which trace more distant, elliptical orbits at an angle to the equator. Usually the regular satellites orbit in the same (‘prograde’) direction as the planet’s rotation, and tend to be bigger. For these reasons, the nebula hypothesis has them forming from the same rotating disc of gas that gathered round the future planet’s rocky core – an explanation that of course depends on the nebula hypothesis.

Irregular moons mostly have a retrograde motion because elliptical orbits are more stable when retrograde. However, Saturn’s irregulars, which also travel on elliptical orbits, include a high proportion of prograde satellites. It is possible to model the capture of both types (Astakhov et al 2003). Size is very variable. At 2,700 km diameter, Triton is by far the biggest of Neptune’s satellites, has an inclined but circular orbit, and lies between a group of small inner prograde satellites and several exterior satellites with both prograde and retrograde orbits. Density calculations suggest that it consists 30-45% of water ice and 55-70% of rocky material.

The present solar system offers no mechanism by which moons could be captured, so the thinking is that captures must have taken place early in its history, when there was a much greater likelihood of collisions and near misses (Jewitt et al 2006).