| Abstract: | Abstract— Crystal‐bearing lunar spherules (CLSs) in lunar breccia (14313, 14315, 14318), soil (68001, 24105), and impact‐melt rock (62295) samples can be classified into two types: feldspathic and olivine‐rich. Feldspathic CLSs contain equant, tabular, or acicular plagioclase grains set in glass or a pyroxene‐olivine mesostasis; the less common olivine‐rich CLSs contain euhedral or skeletal olivine set in glass, or possess a barred‐olivine texture. Bulk‐chemical and mineral‐chemical data strongly suggest that feldspathic CLSs formed by impact melting of mixtures of ferroan anorthosite and Mg‐suite rocks that compose the feldspathic crust of the Moon. It is probable that olivine‐rich CLSs also formed by impact melting, but some appear to have been derived from distinctively magnesian lunar materials, atypical of the Moon's crust. Some CLSs contain reversely‐zoned “relic” plagioclase grains that were not entirely melted during CLS formation, thin (≤5 μm thick) rims of troilite or phosphate, and chemical gradients in glassy mesostases attributed to metasomatism in a volatile‐rich (Na‐K‐P‐rich) environment. Crystal‐bearing lunar spherules were rimmed and metasomatized prior to brecciation. Compound CLS objects are also present; these formed by low‐velocity collisions in an environment, probably an ejecta plume, that contained numerous melt droplets. Factors other than composition were responsible for producing the crystallinity of the CLSs. We agree with previous workers that relatively slow cooling rates and long ballistic travel times were critical features that enabled these impact‐melt droplets to partially or completely crystallize in free‐flight. Moreover, incomplete melting of precursor materials formed nucleation sites that aided subsequent crystallization. Clearly, CLSs do not resemble meteoritic chondrules in all ways. The two types of objects had different precursors and did not experience identical rimming processes, and vapor fractionation appears to have played a less important role in establishing the compositions of CLSs than of chondrules. However, the many detailed similarities between CLSs and chondrules indicate that it is more difficult to rule out an origin for some chondrules by impact melting than some have previously argued. Differences between CLSs, chondrules, and their host rocks possibly can be reconciled with an impact‐melt origin for some chondrules when different precursors, the higher gravity of the Moon compared to chondrite parent bodies, and the likely presence of nebular gas during chondrule formation are taken into account. |
