ON THE REDSHIFT APPARENT SIZE DIAGRAM OF DOUBLE RADIO-SOURCES

We review the data on the angular sizes of double radio sources. A list of 540 Fanaroff-Riley type II double sources is collected from literature, and it is used for a critical discussion of previous work based on smaller samples. We find that results from all previous samples agree well with each other and with our new sample, if the discussion is limited only to well-defined two-sided (''classical') doubles. Different results which have appeared in literature are traced to different sample definitions of double radio sources. In addition, we find that the behavior of high radio power quasars and radio galaxies is different from low radio power radio galaxies which, ff unnoticed, leads to contradicting conclusions. The main conclusions, relevant to all samples studied, are the following: 1. There is a positive correlation between the true radio source size and radio power among low-luminosity radio galaxies. This may be partly explained through sample selection effects. 2. There is a negative correlation between radio size and power among high luminosity radio galaxies and all quasars. It follows the constant total energy envelope closely. 3. There is no significant difference between the radio sizes of radio galaxies and quasars of the same luminosity. However, it is possible that real differences in the radio sizes of the two populations are hidden by selection effects in our sample. 4. The angular size-redshift diagram shows a deficiency of large radio sources at high redshifts which is fully explained by the above-mentioned negative correlation without need for cosmic evolution of radio source size. However, the possibility of some cosmic evolution is not totally ruled out by the data.