Garrington, Conway & Leahy (1991; hereafter GCL) extended the study to a larger sample, confirming the effect and finding in addition asymmetries in spectral index and hotspot brightness. Liu & Pooley (1991) also found the more depolarized side of a source to have a steeper spectrum. These asymmetries are surprising because if the source is intrinsically two-sided then no intrinsic differences between the two sides of the source are expected.
In addition to these asymmetries, the two sides of the source can be different sizes, the source can be bent, and the two sides can be located at different distances from the nucleus. Associated optical line emission is also asymmetric (McCarthy, van Breugel & Kapahi 1991). It is not presently clear how (and whether) these structural asymmetries are related to the depolarization and spectral asymmetries.
In Section 2 I consider mechanisms other than simple orientation which might lead to the observed depolarization asymmetry. I model the depolarization and its asymmetry in Section 3. I fit the polarization curves for a small sample with functional forms suggested by previous work. I attempt to derive or at least constrain the orientation angles of depolarized radio sources. In Section 4 I show that the spectral index asymmetry can be understood as a simple Doppler effect, and that this interpretation of the asymmetry as an orientation effect is supported by its strong correlation with the depolarization asymmetry. I consider hotspot asymmetries in Section 5, and present my conclusions in Section 6. I use H_0 = 50 km/s/Mpc throughout.