Far-Off Planet Makes a Comeback

James Glanz

"It looks like we're back to an ugly old planet," says David Gray. Nearly a year ago, the University of Western Ontario astronomer had issued a serious challenge to the case for an extrasolar planet--the first to be discovered around a sunlike star. He presented evidence that slow jitters in the spectrum of the parent star, thought to result from a planet's periodic tug, were actually due to a pulsation of the star's gases (Nature, 27 February 1997). The ensuing debate was at times less than civil. But harmony has been restored: Gray and at least three other groups say they have been unable to reproduce his earlier results. What looked like a planet killer may have been just a chance alignment of noisy data points.

One group, led by Artie Hatzes of the University of Texas, Austin, published its results in yesterday's issue of Nature alongside Gray's concession. A paper by another group, led by Timothy Brown at the High Altitude Observatory in Boulder, Colorado, has been accepted at Astrophysical Journal Letters. And a third group, at the University of Paris, is still finishing its analysis but also sees "no indication" of pulsations of the star, 51 Pegasi, says team member Jean Schneider. "The only reasonable explanation for the velocity wobble is [still] a planet," concludes Didier Queloz, who made the discovery with Michel Mayor at the Geneva Observatory (Science, 20 October 1995, p.375).

Mayor and Queloz, who is now at the Jet Propulsion Laboratory in Pasadena, California, had monitored hundreds of dark absorption spikes carved into 51 Peg's spectrum by elements in the star's atmosphere such as iron and calcium, which filter out specific frequencies of light. In light from a stationary star, the frequencies would remain fixed. But the observed frequencies shifted up and down by small amounts over a 4.23-day period. Mayor and Queloz inferred that a roughly Jupiter-size planet was whipping around 51 Peg in an orbit much closer than Mercury's around the sun, causing the entire star to wobble and shifting the spike frequencies by the Doppler or train-whistle effect.

Then came Gray's challenge. Using a spectrograph most astronomers describe as somewhat outmoded, he monitored a single absorption spike and found that asymmetries in its shape also changed over a 4.23-day period. A simple Doppler shift couldn't cause the distortion, but Gray believed that both the shape changes and the frequency shifts could be due to a bizarre type of "nonradial" oscillation never seen in a sunlike star: a slow sloshing of the star's surface gases. Because his 39 noisy data points were scattered over 7 years, however, Gray and others now emphasize that there was roughly a one-in-300 chance that a spurious 4.23-day signal might show up by chance.

That appears to be what happened. "I have to conclude that nature played a dirty trick on him," says Brown. In the most conclusive of the new measurements, Hatzes and co-workers made about 120 measurements of several absorption lines at more than twice Gray's spectral resolution and saw no changes in the line shapes. Neither did Gray when he made further observations with his original apparatus. His earlier results, he says, were either a fluke or a transient phenomenon that has since stopped.

But as the technical disagreement fades, the dispute has thrown light on the underside of a high-stakes field where new claims are followed like sports scores by the wider public. Astronomers grumbled privately about the attacks on Gray's work that appeared on an elaborate Web site --complete with links to corporate sponsors--maintained by the planet searcher Geoff Marcy of San Francisco State University. Gray responded in kind on his own Web site, calling some objections "arguments of ignorance." Tempers have since cooled, but Hatzes says: "I didn't like to see that. It should have been a more civilized debate."