Unification scheme of radio galaxies and quasars falsified by their observed size distributions
In the currently popular orientation-based unified scheme, a radio galaxy appears as a quasar when its principal radio-axis happens to be oriented within a certain cone opening angle around the observer's line of sight. Due to geometrical projection, the observed sizes of quasars should therefore appear smaller than those of radio galaxies. We show that this simple, unambiguous prediction of the unified scheme is not borne out by the actually observed angular sizes of radio galaxies and quasars. Except in the original 3CR sample, based on which the unified scheme was proposed, in other much larger samples no statistically significant difference is apparent in the size distributions of radio galaxies and quasars. The population of low-excitation radio galaxies with apparently no hidden quasars inside, which might explain the observed excess number of radio galaxies at low redshifts, cannot still account for the absence of any foreshortening of the sizes of quasars at large redshifts. On the other hand from infrared and X-ray studies there is evidence of hidden quasar within a dusty torus in many RGs, at $z>0.5$. It seems difficult how to reconcile this with the absence of foreshortening of quasar sizes at even these redshifts, and perhaps one has to allow that the major radio axis may not have anything to do with the optical axis of the torus. Otherwise to resolve the dichotomy of radio galaxies and quasars, a scheme quite different from the present might be required.