1958: not a bad year for ozone.

Over the last decade, atmospheric scientists have gathered more than enough evidence to convict chlorofluorocarbons (CFCs) of creating the ozone hole over Antarctica. But that hasn't kept skeptics from trying to launch an appeal.

One point they frequently raise concerns a sketchy report of extremely low ozone readings over Antarctica in 1958, well before CFCs reached dangerous concentrations in the atmosphere. If accurate, such measurements would exonerate CFCs and recast the ozone hole as a natural phenomenon.

Although atmospheric scientists have long since written off these early low-ozone reports, researcher Paul A. Newman of NASA's Goddard Space Flight Center in Greenbelt, Md., decided to look at the 1950s data to settle the issue. In the April 22 SCIENCE, he concludes that "there is no credible evidence for an ozone hole in 1958."

The original readings came from French researchers who participated in the International Geophysical Year in 1958 by performing experiments at the Dumont d'Urville station on the Antarctic coast. For one of their studies, they exposed spectrographic plates to light from the moon, sun, and stars outside the solar system in order to gauge the amount of ultraviolet radiation reaching Earth's surface. Because atmospheric ozone blocks incoming ultraviolet light, the French researchers used their data to infer the amount of ozone in the sky. They recorded concentrations as low as 110 Dobson units (DU) in September and 120 DU in October 1958.

Newman, however, reports that the Dumont d'Urville readings don't square with measurements taken at three other Antarctic sites and a fourth on a nearby island. Ozone amounts dipped no lower than 293 DU in October at these sites. Researchers there took measurements with the Dobson spectrophotometer, a reliable and easy-to-operate tool still in use.

Newman regards the Dumont d'Urville readings as unreliable, for several reasons. Ozone readings at this site varied much more from day to day than did measurements at other locations in 1958. Furthermore, satellite measurements of the Dumont d'Urville area going back to 1978 suggest that the French data from 1958 do not match the typical ozone patterns observed in the region.