1998A&A...331..901S

New observations of head-tail radio galaxies in the Perseus cluster at 21, 49 and 92 cm using the WSRT are presented. Two new head-tail galaxies were discovered bringing the total for the cluster to five. We show total intensity maps for all of them and spectral index distribution and polarization maps for NGC 1265 and IC 310. The observations with high dynamic range, obtained with the WSRT and the redundancy and self-calibration techniques, reveal a large extension of the tail of NGC 1265 both at 49cm and at 92cm. The projected tail bends over an angle of almost 360°. There are several clear differences between the bright and the faint part of the radio tail: within a distance of one or two resolution beams the intensity decreases by more than an order of magnitude and the spectral index steepens from -1 to -2. The simplest model for the tail is that it delineates the orbit of the galaxy through the cluster. Because the tail is seen in projection, the faint part of the tail is visible only when it is not superimposed onto the bright part of the tail. It is likely, however, that large scale motions of the hot intracluster medium also help shape the tail. In this respect it is interesting to note that the tails of both NGC 1265 and IC 310, as well as the low brightness emission of NGC 1275 appear to have their faint extensions shifted to the east. The very long tail of NGC 1265 makes it possible to test at low frequencies models of the ageing process of the radiating electrons. In the faint part of the tail the brightness temperature is very low, which either indicates a very weak magnetic field or a large deviation from equipartition between particle and magnetic energy. The most remarkable properties of the faint part of the tail of NGC1265, namely its constant surface brightness and spectral index, are compared with three synchrotron ageing models, making some standard assumptions. All these models disagree with either the data or one of the assumptions. The Kardashev-Pacholczyk model and the Komissarov-Gubanov model require magnetic fields stronger than the equipartition value; the Jaffe-Perola model and the Komissarov-Gubanov model require an additional process, like in situ reacceleration or bulk streaming motions of electrons in the tail, far in excess of a plausible galaxy speed, to explain the observed spectra.