Cooling and the longevity of polar rings

Many early-type galaxies contain gas in a nearly polar ring. In static potentials the only possible equilibrium configurations for dissipative gas are closed nonintersecting orbits in a preferred plane of the potential. The application of equilibrium models to observed gas velocity fields, however, has proved problematic, suggesting the consideration of a wider range of models including time dependence. As an example of such a configuration we consider gas on a nearly polar orbit in an oblate, rigid potential. We argue that the rapid evolution of highly inclined rings, observed in previous numerical experiments, is due to both artificial instabilities caused by the lack of radiative cooling and due to the lack of numerical resolution. In static oblate and near-oblate potentials, highly inclined, dissipative rings are long-lived and will develop a stable steady state warp, even without self-gravity. The inclusion of radiative cooling, a physical mechanism that is known to act under these circumstances, disposes of the need to invoke either tumbling potentials or supermassive gas rings for stability. For gas at high inclinations the time scales to "find" these equilibria are much shorter than the settling time scale; this casts severe doubts on the validity of the assumption that gas must always move on closed orbits in a preferred plane, often made to infer the structure of the potential well from gas kinematics.