American Scientist

Future generations of astronomers will probably see the 1990s as a watershed. Before that time, the meaning of the word "planet" seemed obvious, even trivial, and completely secure. The discovery of many small bodies orbiting beyond Neptune posed a minor difficulty after 1992, because it raised the question of whether Pluto was really a planet or just another "Kuiper Belt object." And since 1995, the detection of distant worlds revolving around various sunlike stars has proved a real challenge to the long-held categorizations. Which of these bodies are true planets and which are just small partners in binary star systems? Should the definition hinge solely on the mass of the object, or should it also depend on the nature of its orbit? As if these questions were not difficult enough to answer, astronomers must now grapple with a new problem in celestial taxonomy, because some of the extrasolar planets recently discovered are linked to no star at all.

The detection of such "free-floating" planets was anticipated five years ago, when Adam Burrows of the University of Arizona and four colleagues published a paper in Nature that examined the possibility of resolving giant extrasolar planets with modern telescopes. And prospects for the detection of free-floating planets became truly significant in 1996 and 1997, when the first few brown-dwarf stars were reported.

Brown dwarfs are, in essence, failed suns—ones so small that the nuclear fires that power most stars cannot take hold. Astrophysicists believe that the threshold for true stardom lies at about 80 times the mass of Jupiter. A smaller body may briefly burn deuterium (the heavy and more easily fused isotope of hydrogen), but the hydrogen-fusion reactions of normal stars cannot begin easily, and if they do happen, they don't last long. So an object with a mass less than 80 Jupiters is usually labeled a brown-dwarf star.

Fertile grounds for the discovery of brown-dwarf stars are the active star-forming regions. Because the brown dwarfs created in these places are still relatively young and hot, they give off enough infrared light to reveal their presence, despite their diminutive size. Astronomers have now charted scores of brown dwarfs (or brown-dwarf candidates) in such stellar nurseries. Curiously, some of these objects appear to have about the same mass as the larger planets recently discovered around various distant stars. Maria Rosa Zapatero-Osorio (of the Instituto de Astrofísica de Canarias in Tenerife) and several colleagues have determined, for example, that S Ori 47, a small body in the σ Orionus cluster, has a mass of only about 10 to 20 Jupiters. Because this range spans the threshold below which even deuterium fusion fizzles (about a dozen Jupiter masses), the title of their yet-unpublished paper speaks of "reaching the mass boundary between brown dwarfs and giant planets."

Another team, made up of Philip W. Lucas (of the University of Hertfordshire) and Patrick F. Roche (of the University of Oxford) have also searched a stellar nursery in Orion for possible brown dwarfs. They estimate that 13 of their many sightings in Orion's Trapezium cluster are too tiny even to burn deuterium, a conclusion that prompted them to claim discovery of "the first free-floating objects of planetary mass" in the preprint of a paper now in press with the Monthly Notices of the Royal Astronomical Society.

Some would object to the suggestion that such isolated objects are really planets. Lynne A. Hillenbrand, for one, an astronomer at the California Institute of Technology, argues against a definition based solely on mass: "If you're going to call something a planet, you should be sure it formed like a planet." But her Caltech colleague Eduardo L. Martín disagrees. He admits the nomenclature that he and some other astronomers are using for these enigmatic objects might seem bizarre: "This is all very strange ... they are not associated with any star, and still we're calling them planets." But he argues that the classic definition, which depends on how the object formed, involves mechanisms that are as yet poorly understood, whereas the deuterium-burning limit provides a convenient way to distinguish giant planets from brown-dwarf stars according to well-established principles of nuclear physics.

Curiously, neither Hillenbrand nor Martín is willing to accept the British claim of finding free-floating objects of planetary mass. Hillenbrand, who has also probed Orion for such objects, suspects that many of Lucas and Roche's detections are stars outside the cluster masquerading as brown dwarfs or giant planets. "I'm content with the conclusion that they are unrelated objects," she notes. Martín, who is part of the group that found S Ori 47, puts his skepticism more bluntly, pointing out that Lucas and Roche did not support their assertion with spectral evidence that these dim objects are not, in fact, ordinary stars: "If you don't have a spectrum, you don't know what you're talking about; it's as simple as that."

Lucas is, however, sticking to his guns. "I'm 99 percent sure that these objects we're claiming to be planets are planets." His confidence stems, in part, from the spectral observations he gathered since submitting his paper, spectra that he says show just the sort of features one would expect to see coming from giant planets. He has, however, expressed his intent to correct the paper's assertion that he and Roche made the first detection of free-floating objects of planetary mass. It seems that they—along with many science journalists—had completely overlooked a 1998 Science article by a group of Japanese astronomers who had already reported the discovery of two free-floating objects of planetary mass.

In that paper and one that followed shortly afterward (in a similarly prominent publication, the Astrophysical Journal), Motohide Tamura of the National Astronomical Observatory of Japan, along with several Japanese colleagues, described their observations of dim bodies in the Chamaeleon cluster, another site of active star formation. Their estimates for the mass of two isolated objects in the cluster lie below the threshold for burning deuterium, qualifying these objects as free-floating planets, at least according to one definition.

Why didn't the popular press cover this rather momentous discovery? And why had Lucas and Roche failed to realize until now that their detection of free-floating planets was not the first to be reported in the scientific literature? The answer turns out to be quite simple: "The Japanese are very modest, " Lucas explains, "the discovery is not mentioned in the abstract or title of either paper."—David Schneider