We study the effect of the stochastic character of supernova explosions on the anisotropy of galactic cosmic rays below the knee. We conclude that if the bulk of cosmic rays are produced in supernova explosions the observed small and nearly energy independent amplitude of the anisotropy and its phase are to the large extent determined by the history of these explosions in the vicinity of the solar system, namely by the location and the age of the supernova remnants, within a few kpc, which give the highest contibution to the total intensity at the present epoch. Among the most important factors which result in the small magnitude and the energy independence of the anisotropy amplitude are the mixed primary mass composition, the effect of the Single Source and the Galactic Halo. Special attention is given to the phase of the anisotropy. It is shown that the excessive cosmic ray flux from the Outer Galaxy can be due to the location of the Solar System at the inner edge of the Orion Arm which has the enhanced density and rate of supernova explosions.

The advent of new and improved extensive air shower (EAS) arrays - and the attendant generation of data of higher statistical precision than hitherto - has led us to return to an old problem: the origin of the change in spectral slope at an energy of eV. The data recorded are the so-called shower size spectra at ground level, i.e. the spectrum of the total number of cosmic rays inferred for each shower, and we use results from seven arrays which are situated at levels ranging from near sea level to half way up the atmosphere (in terms of atmospheric mass per unit area). The method adopted by us is a new one which enables the combination of experimental data in such a way that previously unrecognized features are now apparent. We present evidence for a spectral shape of the shower size spectrum, and thus the primary particle energy spectrum, which has the signature of an extra component from a single source which protrudes above the `background' due to many other cosmic ray sources in the Galaxy. It is proposed that the `single source' is a local supernova remnant.

We present all-particle primary cosmic-ray energy spectrum in the 3 · 10 6 − 2 · 10 8 GeV energy range obtained by a multi-parametric event-by-event evaluation of the primary energy. The results are obtained on the basis of an expanded EAS data set detected at mountain level (700 g/cm 2 ) by the GAMMA experiment. The energy evaluation method has been developed using the EAS simulation with the SIBYLL interaction model taking into account the response of GAMMA detectors and reconstruction uncertainties of EAS parameters. Nearly unbiased (< 5%) energy estimations regardless of a primary nuclear mass with an accuracy of about 15−10% in the 3 · 10 6 − 2 · 10 8 GeV energy range respectively are attained.An irregularity ('bump') in the spectrum is observed at primary energies of ∼ 7.4 · 10 7 GeV. This bump exceeds a smooth power-law fit to the data by about 4 standard deviations. Not rejecting stochastic nature of the bump completely, we examined the systematic uncertainties of our methods and conclude that they cannot be responsible for the observed feature.

A Fourier analysis of the magnitudes and timing of the Phanerozoic mass extinctions (MEs) demonstrates that many of the periodicities claimed in other analyses are not statistically significant. Moreover we show that the periodicities associated with oscillations of the Solar System about the galactic plane are too irregular to give narrow peaks in the Fourier periodograms. This leads us to conclude that, apart from possibly a small number of major events, astronomical causes for MEs can largely be ruled out.