BALTIMORE—The generic line on dark matter is that nobody really knows what it is because nobody has seen it. The former claim remains basically unassailable—there are many forms dark matter could take. But one research group would dispute the latter assertion. Over the past several years, the Italian DAMA (for DArk MAtter) collaboration has been making the claim that their subterranean detector has registered the signature of dark matter as Earth passes through a sea of the stuff. But despite an ever-strengthening observational case for their claim, the DAMA collaboration's finding remains a source of broad skepticism within the scientific community.

The situation was highlighted this week at a symposium on dark matter here at the Space Telescope Science Institute (STScI). A May 2 talk by a DAMA scientist was immediately followed by a presentation about a competing dark matter detector—one whose results are more widely accepted and whose data contradict DAMA's finding. A few days later a third scientist representing a third detector weighed in, and confused the situation even more.

Dark matter provides the universe with a great deal of its total mass, so it is critical to cosmic evolution, but it is both invisible and barely interactive with normal matter, making it an incredible challenge to detect directly. So far its existence has been inferred from its gravitational effects in shaping galaxies and other large-scale structures.

Pierluigi Belli of the University of Rome Tor Vergata and the DAMA collaboration explained that his team's detector has measured an annual fluctuation in particulate hits that seems to fit the bill for dark matter. DAMA is a detector meant to measure changes in the ambient particle environment—including, presumably, dark matter—as Earth passes through its orbit. "The velocity of Earth in the galactic frame is different at different times of the year," Belli explained. As the sun moves steadily through its orbit around the galaxy, Earth orbits the sun in turn, and the planet's velocity either adds to or subtracts from the sun's velocity. If dark matter rings the galaxy as theory predicts, Earth should be oscillating back and forth through a sea of particles. And DAMA should be able to register a yearlong ebb-and-flow cycle in the number of dark-matter particles passing through the detector.

The DAMA experiment consists of 250 kilograms of crystalline sodium iodide concealed deep underground in Italy's Gran Sasso National Laboratory. Burying the detector underground shields the instrument from more mundane particles, whereas dark matter should pass through rock and into the lab relatively easily. The hope is that some of those particles, having cruised cleanly through Earth, will then bump into one of the atoms in the DAMA detector, announcing its presence by depositing a tiny bit of energy in the crystal.

In more than a dozen years of operation, DAMA has registered a seasonal fluctuation in particle hits that agrees with what a dark matter sea should look like. As predicted, the fluctuation cycle peaks around the start of June and lasts almost exactly a year. "The results are well compatible with many dark matter scenarios," Belli said. Specifically, DAMA could be seeing a very lightweight form of the preferred candidate for dark matter, known as the weakly interacting massive particle, or WIMP. "We have positive evidence for the presence of dark matter particles at a very high confidence level," Belli said.

The general criteria for announcing evidence suggestive of a new particle or physical effect is three standard deviations, or 3 sigma; the benchmark for claiming a new discovery is 5 sigma. The DAMA seasonal flux is now a roughly 9-sigma effect. But doubts of the dark-matter interpretation still loom large.

"I think everyone would agree at this point that they see a signal," says STScI astronomer Mario Livio. "The question is: What is it?" Other researchers at the meeting echoed the same sentiment—after all, a 9-sigma result demands some explanation, even if it is not the dark-matter explanation the experimenters have offered. "It's an intriguing hint," Stanford University physicist Peter Graham said in an April 4 panel discussion at the symposium. "They clearly have a signal, and they've been seeing it for a long time."

Albert de Roeck, a physicist working at the Large Hadron Collider outside Geneva, Switzerland, suspects that DAMA scientists may have gotten off on the wrong foot with their peers. DAMA first announced evidence for a seasonal modulation indicative of dark matter more than 10 years ago, when the signal was much weaker than it is now. Not everyone was convinced. "It sort of seemed like they wanted it to be there," de Roeck says. Now, even with much stronger evidence, the field remains dubious.

Physicists have complained that the DAMA group has not been open enough with its data. During Belli's talk, physicist David Cline of the University of California, Los Angeles, noted that other researchers have not been given access to the raw, unprocessed data from DAMA to see how the collaboration arrived at their 9-sigma result. "It would be nice to get a look at that," Cline said. "Usually in science people do open their doors to let other people look at their data."

Belli said that that was simply not the procedure. "These data have been analyzed in the simple standard method," he said. "So we don't think that it's important to give the particular data."

The neighboring Xenon100 experiment has not helped DAMA's case. Xenon100, which is located in the same underground lab, works in a similar way to DAMA. Its detector, with a 62-kilogram liquid xenon target, has been awaiting collisions from dark matter since 2009. It boasts a very low level of background noise that should make it exquisitely sensitive to detecting ambient dark matter particles.

In a talk immediately following Belli's, Elena Aprile of Columbia University showed data from an April Xenon100 study that appeared to rule out the kind of lightweight dark matter particle favored by DAMA. What is more, the experiment did not pick up any kind of seasonally varying signal. In 100 days of taking data Xenon100 was unable to gather any evidence for the presence of dark matter, despite its sensitivity. "I didn't want to say it in words," Aprile said when asked to explain the discrepancy between her experiment and DAMA, "but…of course, already since some time, these data and others are in friction with that. They're not compatible."

Later, during the panel discussion, Belli was pressed to offer an explanation for what the DAMA signal might represent if it is someday proved not to be dark matter, as most observers suspect. "It is quite difficult to demonstrate that it's not dark matter," Belli said, adding that he and his colleagues have tried to exclude every other effect they can think of. "We cannot explain it in another way."

The last speaker of the symposium gave DAMA a glimmer of hope. Juan Collar of the University of Chicago, presenting May 5 on behalf of the CoGeNT (Coherent Germanium Neutrino Technology) experimental collaboration, announced that his group also registered a faint seasonal modulation in particulate hits. The CoGeNT dark matter detector had its experimental run cut short at 15 months due to a March fire in the Soudan Mine in Minnesota where it resides. But in analyzing that limited data Collar and his colleagues found a possible signature of seasonal modulation akin to what DAMA has seen.

CoGeNT's results are in broad agreement with a DAMA-like dark matter ebb and flow in terms of the timing, duration and amplitude of the seasonal variation. With a few basic assumptions, the CoGeNT signal becomes a 2.8-sigma result—not enough to make a strong statement, but enough to raise a few eyebrows. "We're not going to claim that we're seeing WIMPs," Collar cautioned.

He charged that the Xenon100 collaboration has overstepped in using its own nondetection to rule out the DAMA claim. "Xenon is not a good medium for light WIMPs," Collar said. "Light WIMPs are wicked.... With present technologies there's very little we can say [about them] with certainty." He had harsh words for a smaller, earlier iteration of the Xenon100 detector known as Xenon10. In one of the slides accompanying his talk, Collar referred to some results from Xenon10 as "pure, weapons-grade balonium."

CoGeNT may turn out to be the ally DAMA has long lacked—Belli, who tucked himself quietly in a back corner of the auditorium for most of the symposium, moved up to the third row for the CoGeNT talk. But Collar maintains he is not taking sides. "Maybe DAMA's wrong, maybe they're right, but we have to remain neutral," Collar said. "I find myself caught between the believers and heathens."