Astronomers are uncovering newfound planets in orbit around other stars at a meteoric rate these days. The tally of known planets outside the solar system now stands at more than 450, of which about 50 have been discovered just this year. That pace promises to increase as NASA's Kepler mission carries out its multiyear survey of a large patch of stars; the campaign has already located several hundred planetary candidates for follow-up study and confirmation.

At the same time that missions such as Kepler are hard at work using proved detection methods, researchers are looking to expand the exoplanetary tool kit by demonstrating new ways of locating distant worlds. One method that has been discussed for years but has yet to bear fruit is known as transit timing—if a planet passes in front of its host star so that it blocks out a small but detectable fraction of the star's light, researchers can time the arrival of that partial eclipse, known as a planetary transit. An ordinary, unperturbed orbit will bring the planet around at regular intervals, but the presence of an additional unseen planet can disrupt an exoplanet's orbit, leading to slight variations in the timing of its transit.

In a new study a team of researchers in Germany, Poland and Bulgaria has used transit timing of a known massive exoplanet to identify a hypothetical, much smaller companion. If the hypothetical planet could be confirmed—by no means a sure thing—it would be the first exoplanet to be located by variations in transit timing. The paper, which is available at the online preprint repository, has been accepted for publication in the Monthly Notices of the Royal Astronomical Society.

The researchers timed the transits of WASP 3 b, a massive planet discovered in 2007 that orbits a star larger than the sun some 700 light-years away. (The planet and its host star, known as WASP 3, are named for the search program that found the planet, the Wide-Angle Search for Planets.) WASP 3 b passes in front of its star every 44 hours, 19 minutes and 26.5 seconds, on average, but the study's authors found that the actual arrival time of individual transits deviated from predictions by up to about three minutes.

"We observed those transits and we noticed that some transits occur a little bit earlier or a little bit later than theoretical predictions," says lead study author Gracjan Maciejewski, an astronomer at the Friedrich Schiller University of Jena in Germany. "That was a hint for us that there must be something strange in the system."

Maciejewski and his colleagues modeled the various planetary systems that could explain the varied transit times. "We found that there is one configuration that describes all those things that we observe in the best way," he says. The researchers concluded that the most likely perturber of WASP 3 b's orbit is an additional planet, about 15 times the mass of Earth, following a wider orbit than that of the known planet.

But inferring a new world's existence with a handful of measurements is much different than pinning down a planet's properties with a variety of complementary techniques. "You can see that this is an indirect method," Maciejewski says, acknowledging that others have claimed planetary detections by transit timing in the past, only to have other astronomers rule out the purported finds.

And indeed, some researchers in the field remain unconvinced by the new research. "Unfortunately, my opinion is basically that it is another candidate that will need confirming rather than a solid result that you can hang your hat on," says Jacob Bean, an astronomer at the Harvard–Smithsonian Center for Astrophysics. Estimating the errors inherent in the kind of measurements used by Maciejewski and his co-authors is a tough thing to do, Bean says; if the actual errors in the data were larger than the researchers had accounted for, the variations in the observed transit times could vanish. And even if the accounting is solid, he adds, a 15-Earth-mass planet may not be the cause of the orbital disruption. "They don't have nearly enough data points to identify a unique solution," he says. "So it could be a planet like the one they propose, but I suspect there are other possible solutions as well."

One of the early proposers of using transit timing to discover new worlds remains similarly skeptical. "I'm not too hopeful that this is a planet," says astronomer Eric Agol of the University of Washington in Seattle. Given that transit timing is an unproved method for discovering exoplanets, Agol would like to see a signal-to-noise ratio much higher than that found by the European group as well as confirmation by another observational technique. "Maybe more data will turn something up—maybe I'm wrong," Agol says. "Mainly, this is a study that calls for more study."

Maciejewski says that he and his colleagues are busy taking new observations of the system, and that final results from this year's campaign should be in hand by early October.