PDF《**FULL TITLE**》

3.5 and b/a >

0.6. We draw attention to the fact that we are dealing with relatively round galaxies. Though the major axis alignment for the spiral galaxies is interesting in its own right, for the rest of this contribution we will concentrate our discussion on the early-type galaxies. An extensive discussion of the behaviour of both types of galaxy can be found in Battye &

Browne (2009). The bias towards ?α = 90? in the red objects with high concentrations is highly statistically signi?cant. Since in these predominantly elliptical galax- ies the radio-emission is almost certainly powered by mass accretion on to a central disk/black hole system, the radio elongation we measure will be that of the overall spin axis of that system. Our results con?rm those reported by Condon et al. (1991) who saw a strong trend for the jet axes in UGC ellip- tical galaxies to be aligned with the optical minor axes. Simulations done by Sansom et al. (1987) for di?erent galaxy geometries show that strong minor axis alignments are observable only for galaxies which are oblate spheroids. There- fore our results imply that there is a bias for the spin axis of the central engine to be aligned with the minor axes of galaxies and also that these galaxies are predominantly oblate spheroids, something supported by the analysis of galaxy shapes (Padilla &

Strauss (2008)). The radio minor axis alignment is something one might expect to see if the galaxies were rotationally supported and the black hole accretion disk axes were aligned with the overall galaxy rotation axes. One should contrast what we see here in elliptical galaxies with that seen in Seyfert galaxies where there is only a weak relationship between jet axis and the axis of the host galaxy disk (Kinney et al. (2000);

Gallimore et al. (2006);

Raban et al. (2009)). We have divided the sample further by the ratio of radio to optical ?ux density expressed as t ? r where r is the r magnitude and t is given by: t = ?2.5 log10 Sint 3631Jy . (2) What is most remarkable is that the strong ?α = 90? bias in the elliptical galaxies is con?ned to the radio quieter subset of these objects as is clearly evident from Fig.2 The division at t ? r = ?2.5 produces sub-samples with a ratio in numbers of around 2:1 and was not chosen a priori to have any particular physical signi?cance for the elliptical galaxy population. However, the division seems to have physical signi?cance, indicating that there are two distinct types of object within the elliptical population exhibiting quite di?erent behaviour. We speculate that the galaxy shape in the quieter objects is ?xed by rotation and that jets emerge along the stellar rotation axis. We emphasize that the two radio populations we talk about here have nothing to do with the traditional Jet orientations

5 Figure 3. Plot of the stellar masses of the sample split by the ratio of radio to optical luminosity. The lower luminosity objects are indicated by the dashed line and the higher by the continuous line. FR1/FR2 radio morphological division that for an L? galaxy occurs at around t ? r = ?7 not t ? r = ?2.5. Virtually all our galaxies are in the FR1 range. 4. Discussion;

two varieties of elliptical galaxies We ?nd that amongst the radio quieter objects the radio emission is strongly aligned with the optical minor axes. This requires there to be negligible pro- jection e?ects suggestin........