Trapped gas proves a meteorite came from Mars
Gas sealed in shock glass inside a 7.9 kg Antarctic stone matched the air Viking had measured in 1976. Terrestrial laboratories suddenly held free samples of Mars.
On 12 August 1983 Science ran four pages from two geochemists at NASA's Johnson Space Center under a title that kept its question mark: "Martian Gases in an Antarctic Meteorite?" Bogard and Johnson had heated a pod of black shock-melted glass out of EETA79001, a 7,942-gram stone picked off the ice at Elephant Moraine, Antarctica, in 1979. Twenty-odd such pods sit in its sawn faces; the one that mattered was labelled BRAVO. It gave up argon at an argon-40 to argon-36 ratio above 2,000, xenon enriched in xenon-129, krypton and xenon in a ratio near 15. Earth's air sits near 300 on that argon number. Viking, sampling martian air in 1976, had reported roughly 3,000.
These rocks had been an embarrassment for a decade. Shergotty fell in Bihar in 1865, Nakhla near Alexandria in 1911, Chassigny in Haute-Marne in 1815, and by the 1970s dating put the nakhlites near 1.3 billion years and the shergottites closer to 180 million. Almost every other meteorite is 4.5 billion years old. A basalt erupted 180 million years ago needs a parent body still hot enough to melt — a planet. Mars was the only candidate, and the objection looked fatal: escape velocity there is 5.03 km/s, and an impact that violent ought to vaporise rock, not throw it clear.
Shock melt traps whatever gas surrounds it at the instant of impact, and what this glass had swallowed matched no meteorite in any collection. It matched Mars. Becker and Pepin extended the match to nitrogen isotopes in 1984, Ott and Begemann confirmed it in 1985, and by 1987 the literature had dropped the question mark. H. J. Melosh supplied the mechanism: spallation flings near-surface rock off the edge of the shock front fast and barely damaged. EETA79001 crystallised 169 ± 23 million years ago, spent about 600,000 years in space on the cosmic-ray clock, then lay in Antarctic ice for perhaps twelve thousand more.
Overnight, laboratories on Earth held free samples of Mars. NASA's compendium listed 62 named martian meteorites in 2012 and 104 by 2017; the wider catalogue, counting desert fragments separately, has passed two hundred. Much of what we claim about the planet's interior runs through them: the hafnium-tungsten clock on the core, the chemistry that checks seismic models of the mantle. Ten years on the same test was run on a misfiled stone in a Houston cabinet, and ALH84001 became martian.
It is also the standing argument against Mars Sample Return — and the answer to it. A meteorite arrives with no address, no crater, no lava flow, not even a hemisphere, and geology without context is half a measurement. The delivery mechanism biases the set hard toward igneous rock: coherent basalt survives launch and entry, the mudstones Curiosity drills in Gale crater would not. Contamination starts on landing: EETA79001 grew a crust of terrestrial weathering along its glass veins. A sealed tube in Jezero has none of those problems. It just has no ride home.