COSMIC EVOLUTION AND LUMINOSITY DEPENDENCE OF THE PHYSICAL SIZES OF POWERFUL RADIO GALAXIES AND QUASARS

The redshift (z) and luminosity (P) dependence of the physical sizes (l) of powerful radio sources is examined, separately for radio galaxies and quasars. Our sample of 789 sources, which includes recent data that have become available on high-redshift radio galaxies, is sufficiently large to populate significant portions of the P-z plane. The distribution of the median values of observed sizes in different P-z bins seems to be statistically well described by the functional form l is proportional to P(beta)(1 + z)n, both for radio galaxies and, independently, for quasars. It turns out, however, that the behaviour of the luminosity-size correlation is quite different in quasars from that in radio galaxies. Amongst the quasars at any given redshift, the more luminous sources have statistically smaller physical sizes (beta approximately - 0.2), which is a reversal of the trend seen amongst radio galaxies (beta congruent-to + 0.35). Moreover, in a given luminosity range, there are statistically significant differences between the cosmic evolution of quasar sizes and that of radio galaxy sizes. For quasars there is, if anything, only a marginal size evolution (n approximately - 0.2), while the sizes of radio galaxies seem to fall rapidly with redshift (n congruent-to - 3.0). These results support our earlier claim that there appears to be an intrinsic difference between the size distributions of radio galaxies and quasars. Based on these results, the hypothesis of the unified scheme, where the differences in the observed sizes of radio galaxies and quasars are attributed to geometric projection effects arising from their supposedly different orientations with respect to the observer's line of sight, can be rejected above a statistical significance level of 0.01 per cent.