Newcomer #8, also from Super-Kamiokande, was submitted four months earlier than #1 and describes analysis leading researchers to suspect detection of neutrino oscillations. Super-Kamiokande is a 50-kiloton water Cherenkov detector deep below Mt. Ikenoyama, Japan. Over 13,000 photomultiplier tubes detect Cherenkov radiation from interactions of neutrinos with water. In an 18-month period neutrinos triggered the detectors 4 x 10⁸times. Most hits are caused by cosmic ray muons and radioactivity in the detectors. Meticulous sieving of the data produced just 900 muon-like and 983 electron-like detections of atmospheric neutrinos. The next step was to compare the observed ratio of muon-like to electron-like events with a Monte Carlo simulation. Theory predicts unity whereas Super-Kamiokande scored 0.61 for this key ratio. The detector has greater mass and sensitivity than previous experiments so the divergence of theory and experiment is highly significant. In #8 the team suggests that the ratio of muon-like to electron-like neutrinos from atmospheric collisions is smaller than expected. Paper #1 identifies neutrino oscillations as the cause.

   The hot areas of observational cosmology are structure formation (#6) and star formation (#7, #10). Over 25 years ago, observers discovered that galaxies are surrounded by extensive massive haloes of dark matter. Stars and hot gas trace out spiral arms or elliptical profiles, but they account for only 10% of the mass. Elusive cold dark matter (CDM) is the ballast.

   In #6 Julio Navarro (University of Arizona), Carlos Frenk (University of Durham, U.K.) and Simon White (Max Planck Institute for Astrophysics, Garching, Germany) publish density profiles (halo mass density as a function of distance from the galactic center) computed via N-body simulations using a wide variety of assumptions. Their astonishing conclusion is that the profiles have the same shape, independent of the halo mass, the power spectrum of initial fluctuations (which trigger galaxy formation), and the cosmological parameters W and L. Galaxy formation in the early universe was a rapid and violent process, with mergers and collisions playing a large role in molding galaxies. The situation could hardly be simpler: halos of all masses in all hierarchical cosmologies look the same and their densities are proportional to the cosmic density at the time they formed.

   Hughes et al. (#7) and Barger et al. (#10) used the U.K. Submillimetre Common User Bolometer Array (SCUBA), a detector with superb sensitivity at 800 mm. Hughes et al. targeted the Hubble Deep Field (HDF) in a 50-hr integration taken under exceptional observing conditions. They extracted five strong far-IR (infrared) sources, and matched four of them with HDF galaxies at redshifts 2 < z < 4. Optical surveys in the HDF had indicated a peak of star formation at redshifts between 1 and 1.5. SCUBA has overturned that, showing that star formation was a factor of 5 higher much earlier in the history of the universe at redshifts 2 - 4. Barger’s team used SCUBA on two regions of sky for a total integration of 80 hr. At 850 mm they detected two IR sources which, they argue, are galaxies powered by star formation rates of more than 100 solar masses a year. No direct redshifts are available yet for any of these dusty galaxies so it is possible the time of star formation could be pushed back even further.