Host galaxy sizes of powerful radio sources

This paper analyzes the galaxy sizes and shapes of the surface brightness profiles of the host galaxies of low-redshift (0.03 < z < 0.22), powerful radio galaxies (with powers, P-408 MHz, ranging from 10(25) to 10(29) W Hz(-1)). The aim is to analyze how radio galaxies differ from normal radio-quiet elliptical galaxies and to examine the connection between these properties and the radio activity. At the 22nd isophotal radius (where the surface brightness in the V filter band is 22 mag arcsec(-2)) radio galaxies are 11% +/- 1% larger, at the 24th isophote they are 15% +/- 2% larger, and at the 25th isophote they are 20% +/- 3% larger than radio-quiet elliptical galaxies of the same absolute 16 kpc aperture magnitude. F-R I galaxies exhibit a weak correlation between the size of the host galaxy and the radio power. Since it was shown previously in the literature that optically brighter F-R IIs are also more powerful, a correlation between galaxy size and radio power is expected. However, after the removal of the magnitude dependence of the galaxy size, a residual correlation between size and power is observed, which is stronger than that between the actual size and radio power. The residual correlations of radio power with radius (at the 22nd, 24th, and 25th isophote) are significant at the 5 sigma levels. Thus the radio power is correlated to the ''excess size,'' i.e., by how much bigger the host galaxy is compared to a radio-quiet elliptical galaxy of the same magnitude. F-R II radio galaxies do not show such a correlation. A method to compare the profile shapes of radio galaxies to radio-quiet ellipticals is presented, and it is shown that the surface brightness profiles of radio galaxies are relatively noisier. Since 77% of them have a companion within 50 kpc, the profile disturbances are likely to be caused by a galaxy encounter. About twice as many F-R II galaxies have relatively disturbed profiles as F-R I galaxies, and there is an indication that profile disturbances among F-R IIs are more severe. Those F-R IIs that show profile disturbances are on average 4.1 +/- 1.2 times more powerful than those with smooth profiles. F-R I galaxies, on the other hand, do not show such a dependence on radio power. The nature of the correlation between the radio power and galaxy sizes and profiles disturbances are discussed, and together with the results presented in Papers I and II, it is suggested that F-R I sources preferentially form in the most dominant galaxies in dynamically evolved groups and clusters. These galaxies are likely to have been exposed to a relatively large amount of galactic cannibalism, thus giving rise to the larger galaxy sizes. It is speculated that cannibalism may also provide a larger supply of fuel to power the radio engine. F-R II sources, on the other hand, are more likely associated with violent galaxy encounters in dynamically less evolved groups. In fact, galaxy interactions enhance, and probably also trigger, the radio activity in F-R IIs.