Coronal mass ejections (CMEs) are often assumed to be magnetic flux ropes, but direct proof has been lacking.A key feature, resulting from the translational symmetry of a flux rope, is that the total transverse pressure as well as the axial magnetic field has the same functional form over the vector potential along any crossing of the flux rope. We test this feature (and hence the flux-rope structure) by reconstructing the 2007 May 22 magnetic cloud (MC) observed at STEREO B, Wind/ACE, and possibly STEREO A with the Grad-Shafranov (GS) method. The model output from reconstruction at STEREO B agrees fairly well with the magnetic field and thermal pressure observed at ACE/Wind; the separation between STEREO B and ACE/Wind is about 0.06 AU, almost half of the MC radial width. For the first time, we reproduce observations at one spacecraft with data from another wellseparated spacecraft, which provides compelling evidence for the flux-rope geometry and is of importance for understanding CME initiation and propagation. We also discuss the global configuration of the MC at different spacecraft on the basis of the reconstruction results.

We show that temperature anisotropies induced at a shock can account for interplanetary and planetary bow shock observations. Shocked plasma with enhanced plasma b is preferentially unstable to the mirror-mode instability downstream of a quasi-perpendicular shock and to the fire-hose instability downstream of a quasiparallel shock, consistent with magnetic fluctuations observed downstream of a large variety of shocks. Our theoretical analysis of the solar wind termination shock suggests that the magnetic holes observed by Voyager 1 in the heliosheath are produced by the mirror-mode instability. The results are also of astrophysical interest, providing an energy source for plasma heating.