The Viking biology results that never settled
One of Viking's three biology experiments came back positive at both landing sites. The gas chromatograph found no organic molecules at all, and the argument has never closed.
Each Viking lander carried three biology experiments in a box roughly a cubic foot, plus a separate gas chromatograph–mass spectrometer; a robotic arm scooped martian soil into all four. This remains the only direct life-detection experiment flown to another planet.
The Labeled Release experiment, run by Gilbert Levin of Biospherics, dripped a dilute nutrient solution onto the soil, its carbon atoms replaced by radioactive carbon-14: anything that metabolised them would exhale tagged carbon dioxide for the detector to count. At both landing sites, some 6,500 kilometres apart, the counter immediately climbed; a duplicate sample heated to 160°C beforehand gave no response.
The other instruments did not agree. Gas Exchange saw a burst of oxygen the moment the soil was humidified, too much and too fast, without the follow-through biology would give. Pyrolytic Release returned weak signals in seven of nine tests. Klaus Biemann's mass spectrometer, sensitive to parts per billion, found no organic molecules in the soil, and that decided it: meteorites should have salted the surface with carbon compounds for billions of years even without life, so finding none implied something there was destroying organics, a soil aggressive enough to fake the Labeled Release signal. The consensus was chemistry: reactive oxidants built by ultraviolet light in a thin, unshielded atmosphere.
In 2008 NASA's Phoenix lander found perchlorate salts in the arctic soil. Heating perchlorate with organics destroys them and produces chlorinated by-products; Viking baked every sample to hundreds of degrees before measuring it. Chlorinated compounds had appeared in Viking's data, written off as contamination from Earth-side cleaning fluid. In 2018 Curiosity's SAM found organic molecules preserved in three-billion-year-old mudstone at Gale crater.
That does not make the Labeled Release a life detection. Most researchers still favour perchlorate and peroxide chemistry; no abiotic model reproduces every feature of the data, and no biological one is required. Levin argued for life on Mars until his death on 26 July 2021, aged 97. Nobody has flown the experiment again in fifty years.