The first results from a long anticipated experiment have heaped doubt on decade-old observations that hinted at the existence of a fourth type of neutrino, a ghostly particle that rarely interacts with ordinary matter. Researchers from the MiniBooNE experiment at the Fermi National Accelerator Laboratory in Batavia, Ill., announced the finding yesterday.

Confirmation of the fourth neutrino would have given researchers a sign that something was wrong with their highly successful Standard Model, which describes the known particles. "We've all been waiting on pins and needles," says neutrino physicist Edward Kearns of Boston University, who works on other neutrino experiments. Seeing something new is always more exciting, he says, but "it's just extremely important that we get this out of the way and focus on other ideas."

Neutrinos come in three types, or flavors: electron, muon and tau. Experiments conducted over the past decade have confirmed that neutrinos can oscillate back and forth between flavors.

The MiniBooNE team was looking for signs of muon neutrinos morphing into the electron variety.

An earlier experiment, called the Liquid Scintillator Neutrino Detector (LSND) carried out by the Los Alamos National Laboratory (LANL) in New Mexico, found an excess of electron neutrinos after bombarding a target with antimuon neutrinos (the muon neutrino's antimatter counterpart). If flavor change was the culprit, one explanation would be the existence of a fourth, "sterile" neutrino that interacts with matter even less than the other three, all of which continuously stream through Earth unnoticed, occasionally pinging off of an atomic nucleus.

The results were "sort of like an odd duck—they didn't fit," Kearns says.

Tasked to check the finding, MiniBooNE researchers fired a beam of normal matter muon neutrinos at a 12-meter-wide spherical tank of mineral oil 500 meters away. When an electron neutrino strikes a carbon nucleus in the oil, it emits a flash of bluish light. The beam produces electron neutrinos directly at a certain rate; a higher frequency would signal something unanticipated.

The group reported no overall increase in the rate of electron neutrino events for muon neutrinos of different energies, which would have bolstered the LSND case for neutrinos switching flavors. "We have quite conclusively ruled out that model for what might be going on," says Columbia University physicist Janet Conrad, a spokesperson for the MiniBooNE team.

But there are other, more exotic models that still have to be checked, says co-spokesperson William Louis of LANL, who worked on LSND. "That is the $64,000 question,'' he says. "Are the MiniBooNE results consistent with LSND, and is the LSND excess a real signal?'' The team has begun running the experiment with antineutrinos to rule out other possibilities, such as a slight asymmetry between matter and antimatter.

The group did spot an odd uptick in the number of electron neutrinos at lower energies—369 events instead of 273. But Louis says the significance of that is unclear. It could mean that physicists have overlooked a subtle detail in the experiment or miscalculated the rate at which neutrinos collide with atomic nuclei.

Kearns says he suspects a subtle but mundane effect is at play: "It smells to me like something that will wilt under further scrutiny."