Origin, age, and composition of meteorites

Abstract

This paper attempts to bring together and evaluate all significant evidence on the origin of meteorites.

The iron meteorites seem to have formed at low pressures. Laboratory evidence shows that the absence of a Widmanstätten pattern in meteorites with > 16% Ni cannot be attributed to high pressures, but to supercooling or an unusually fast cooling rate for these meteorites, which prevented the development of a pattern. The presence of tridymite in the Steinbach siderophyre provides further, direct proof that the Widmanstätten pattern can form at pressures less than 3 kb. Neither diamond, nor cliftonite, nor cohenite are reliable pressure indicators in meteorites. Diamonds were formed by shock while cliftonite may have been derived from a cubic carbide such as Fe₄C. Cohenite is apparently stabilized by kinetic rather than thermodynamic factors. Several lines of evidence suggest that the irons come from more than one parent body, perhaps as many as four.

The frequency of pallasites is perfectly consistent with an origin in the transition zone between core and mantle of the parent body. “Hybrid” meteorites such as Brenham are not necessarily derived from the metal-silicate interface, but probably resulted from dendrite growth in the solidifying melt.

Ordinary chondrites definitely are equilibrium assemblages rather than chance conglomerates. According to the best available evidence, Prior's rules seem to be valid. The metal particles in chondrites differentiated into kamacite and taenite in their present location, rather than in a remote earlier environment. Trace element abundances in ordinary and carbonaceous chondrites suggest that these meteorites accreted from two types of matter: an undepleted fraction that separated from its complement of gases at low temperatures, and a depleted fraction that lost its gases at high temperatures. These two fractions of primitive meteoritic matter are tentatively identified with the matrix and chondrules-plus-metal, respectively. New restrictive limits are placed on the iron-silicate fractionation in chondrites. No direct evolutionary path exists that connects the currently accepted solar abundances of Fe and Ni and the observed Fe/Si and Ni/Si ratios in chondrites. Apparently the solar abundance of iron is in error. The iron-silicate fractionation seems to have occurred while chondritic matter was in a more strongly reduced state than its present one.

The U-He and K-Ar ages of hypersthene chondrites are systematically shorter than those of bronzite chondrites. Short ages are correlated with shock effects, and it seems that the hypersthene chondrites suffered reheating and partial-to-complete outgassing 0.4 AE ago. The cosmic-ray exposure ages of all classes of meteorites cluster distinctly, indicating that the meteorites were produced in a few discrete major collisions rather than by a quasi-continuum of smaller ones. The dates of the principal breakups are: irons, 0.6 and 0.9 AE; aubrites, 45 m.y.; bronzite chondrites, 4 m.y.; hypersthene chondrites, 0.025, 3, 7–13, and 16–31 m.y. All four clusters of hypersthene chondrites show evidence of severe outgassing 0.4 AE ago, which implies that most or all hypersthene chondrites come from the same parent body.

As already noted by Signer and Suess, two distinct types of primordial gas occur in meteorites. Differentiated meteorites always contain unfractionated gas, while relatively undifferentiated meteorites contain fractionated gas. The former component is invariably associated with shock effects, and seems to have been derived from the solar wind. The latter component is correlated with other volatiles and seems to be a truly primitive constituent of meteoritic matter. Isotopic anomalies in the fractionated gas suggest that meteoritic matter was irradiated with ⩾ 10¹⁷ protons/cm² at a very early stage of its history.

There is very little doubt that most, if not all, meteorites come from the asteroid belt rather than from the moon. The orbits and geocentric velocities of stony meteorites resemble those of the Apollo asteroids (most of which are former members of the asteroid belt that have strayed into terrestrial space), but disagree strongly with the calculated orbits and velocities for lunar ejecta. Öpik's conclusions about the difficulty of accelerating lunar debris to escape velocity represent a further argument against a lunar origin of stony meteorites.

The most likely parent bodies of the meteorites are the 34 asteroids which cross the orbit of Mars. Collisional debris from these objects will remain in Mars-crossing orbits, and perturbations by Mars will inject some fraction of this material into terrestrial space. Most of the Mars asteroids, comprising 98% of the mass and 92% of the cross-section, belong to three Hirayama families (Phocaea, Desiderata, and Aethra), and an additional, previously unrecognized family. These families were apparently produced by disruption of parent asteroids ca. 104, 105, and 46 km in diameter. The size distribution and light curves of asteroids indicate that the larger asteroids are original accretions, rather than collision fragments. There is no reason to believe that the meteorites ever resided in bodies larger than Ceres (d = 770 km).

Various theories on the origin of the meteorites are critically reviewed in the light of the preceding evidence. Wood's theory, which postulates a high-temperature and a low-temperature variety of primordial matter, is in best accord with the evidence. Apparently the asteroids accreted from varying proportions of these two types of material, and were then heated by extinct radioactivity produced in the early irradiation.