1997A&A...323..674E

We test through stellar N-body simulations some scenarios to explain the dynamics of the peculiar nucleus of the Andromeda galaxy (M 31): although HST observations reveal a double nucleus morphology, the rotation field is almost symmetric around the bulge gravity centre and the velocity dispersion is off-centred. We show that any m=1 perturbation has a very short life-time (a few 10⁵yr). Assuming that the bright peak (P1) is a cold stellar cluster infalling into the nucleus, and that the large central velocity gradient is due to a central dark mass (in the range 7x10⁷-10⁸M_☉), we obtain a reasonably good fit to the observations. However, if this cluster lies in the central 20pc, we estimate the life-time of the cluster to be less than 0.5Myr. The dynamical friction is more efficient than estimated by analytic formulae, and is essentially due to the deformation of the stellar cluster through the huge tidal forces provided by the black hole. We show that the cluster cannot be on a circular orbit around the centre if the nucleus hosts a massive black hole of a few 10⁷M_☉, and finally provide some estimates of the kinematics as observed with HST.