Cosmic ray drift, the second knee and galactic anisotropies

We show that the second knee in the cosmic ray spectrum (i.e. the steepening occurring at E ≃ 4× 10 eV) could be related to drift effects affecting the heaviest nuclear component, the iron group nuclei, in a scenario in which the knee at 3×10 eV indicates the onset of drift effects in the lighter proton component. We also study the anisotropies resulting from diffusion and drift currents in the Galaxy, showing their potential relevance to account for the AGASA observations at E ∼ 10 eV, before the extragalactic component becomes dominant. There are a couple of observations regarding cosmic rays (CRs) at energies ∼ 10 eV, i.e. just below the ankle, which still remain puzzling (see e.g. [1, 2]). One is the steepening in the spectrum taking place at an energy Esk ≃ 4× 10 17 eV, which is most apparent in the Yakutsk [3] and Fly’s Eye stereo [4, 5] data, and is usually referred to as the ‘second knee’, at which the spectral index α (such that dN/dE ∝ E) changes from α ≃ 3 below Esk to α ≃ 3.3 above Esk, although probably in a smooth way. The other is the 4% anisotropy at E ≃ 0.8−2×10 eV reported by the AGASA collaboration [6], with an excess observed from a direction near the galactic center and a (probably associated) deficit near the galactic anticenter direction. Also some excess of events in the direction of the Cygnus region (along the Orion spiral arm) has been observed. An excess from a direction near the galactic center was also found in the SUGAR data [7], but it has a smaller spatial extension (consistent with a point-like source) and its direction is few degrees away from the AGASA reported excess. The fact that these anisotropies are associated to galactic features is quite