[en] This paper examines the possibility that magnetospheric processes may, through the direct transfer of magnetospheric energy to the lower atmosphere, be involved in the apparent relationships between solar activity and meteorological phenomena. Empirical evidence for Sun-weather relationships is referred to briefly, but attention is directed primarily towards a discussion of possible direct physical mechanisms that could conceivably link the magnetosphere and the troposphere. Estimates are obtained for the characteristic powers and energies associated with various magnetospheric processes; these are then compared with the power required to drive the normal atmospheric circulation system, the total kinetic energy of this system, and the extra energy required to account for the observed enhancements of vorticity in low-pressure troughs that deepen several days after geomagnetic storms. The efficiency with which magnetospheric convection can be coupled to the troposphere is examined critically and it is argued that the magnetospheric processes are unlikely to produce significant global meteorological changes by any direct mechanism. It is concluded that attention should be focused on a search for an indirect mechanism. (author)