There’s been a flurry of puzzling dark-matter results from astrophysical experiments over the past couple of weeks. Direct-detection experiments on Earth have the potential to give a clearer picture, but in practice they have been barely less contentious.

Cosmological and galaxy observations suggest dark matter makes up more than 80% of the matter in the universe, but particles of it have yet to be definitively detected. Here Nature considers recent hints from astrophysical sources in the context of earlier findings by ground-based experiments, which should be directly sensitive to dark matter passing through the Earth.

Recent controversy began with reports that NASA’s Fermi Gamma-ray Telescope had confirmed a gamma-ray signature that matched an antimatter signal seen by the satellite-borne experiment PAMELA. The antimatter was in the form of positrons (the antimatter counterpart to electrons) and thought to be produced by dark matter particles annihilating in the halo of the galaxy. Unfortunately, the gamma-ray signal wasn't consistent with the dark matter interpretation.

Then, on 29 November, New Scientist suggested that two reanalyses of data from PAMELA (here and here) pointed to dark matter particles that were unexpectedly heavy.

On 1 December, Physics World described a re-interpretation of data that suggetss that the balloon experiment ARCADE may have seen lightweight particles.

That hint that dark matter particles may be at the lighter end of the possible range is consistent with the findings from some, but not all, ground-based experiments. In April, for example, the XENON-100 experiment at Gran Sasso National Laboratory in Italy, came out with the strongest negative limits yet set on lightweight dark matter particles.

Controversy soon erupted when a rival experiment, the Coherent Germanium Neutrino Technology (CoGeNT) experiment in the Soudan mine in Minnesota reported a conflicting positive result,finding particles at the light end of the range that XENON-100 had claimed to rule out. It is worth noting that these particles may be inconsistent with the particles seen by PAMELA, if those are heavy as New Scientist reports.

n addition to reporting its own result, CoGeNT attacked XENON100's limits, suggesting that their detector wasn't sensitive enough to detect particles in the lightweight range that CoGeNT appeared to see something. The flamboyant leader of CoGeNT, physicist Juan Collar of the University of Chicago, ruffled feathers all around with his fierce attacks on the reliability of XENON100's calibrations.