SIMBAD references

We have conducted the largest investigation to date into the origin of phase-resolved, apparent rotation measure (RM) variations in the polarized signals of radio pulsars. From a sample of 98 pulsars based on observations at 1.4 GHz with the Parkes radio telescope, we carefully quantified systematic and statistical errors on the measured RMs. A total of 42 pulsars showed significant phase-resolved RM variations. We show that both magnetospheric and scattering effects can cause these apparent variations. There is a clear correlation between complex profiles and the degree of RM variability, in addition to deviations from the Faraday law. Therefore, we conclude that scattering cannot be the only cause of RM variations, and show clear examples where magnetospheric effects dominate. It is likely that, given sufficient signal to noise, such effects will be present in all radio pulsars. These signatures provide a tool to probe the propagation of the radio emission through the magnetosphere.