Improving the triaxial bulge model of M31

A detailed hydrodynamical model of the gas flow in the triaxial gravitational potential of the bulge of the Andromeda galaxy (M31) has recently been proposed by Berman, and shown to provide excellent agreement with the CO emission-line velocities observed along its major axis. In the present paper, we confirm the validity of that model by showing that it can also reproduce the CO velocities observed off the major axis - a much more robust test. The CO observations, however, tend to span a wider range of velocities than a direct application of the original model of Berman would suggest. This situation can be improved significantly if the molecular disc is made thicker, a requirement already encountered in dynamical simulations of other spiral galaxies, and typically attributed to a broadening of the molecular layer in galactic fountain-like processes. In the central regions of M31, however, it is unclear whether there actually is a thick molecular disc, or whether broadening the molecular layer is merely an artificial theoretical means of accounting for some disc warping. Other effects not included in the model, such as hydraulic jumps, might also contribute to a widening of the range of velocities.