## 9. ConclusionIn this paper we investigated the special rôle of the Coriolis forces in influencing the flow topology over the surface of a contact binary. In order to follow the dynamical effects more easily, the surface geometry was drastically simplified and the mass flow was imagined to be generated by a source-sink pair. We found that, whereas under neglect of Coriolis forces the surface is covered by streamlines connecting source and sink, the inclusion of these forces gave a quite different picture. Most of the surface was found to be covered by closed streamlines and the flow between source and sink was restricted to a narrow equatorial channel running around one side of the object only. Freedom to change the vorticity distribution over the closed streamlines brought us also the freedom to influence the distribution of Jacobi energy over the part of the surface covered by closed streamlines. Hence there is no reason why in any given situation the Jacobi energy should be constant over the surface. This last result assumes that the Coriolis forces have been allowed for. It is not in conflict with the properties of the published models listed in the last section provided this important proviso is observed. An argument, based upon analytical continuation, in favour of the view that Coriolis forces can not be expected to influence the flow topology, is subjected to criticism. Returning now to the Jacobi energy (Bernoulli constant) situation our conclusion can best be summarized by noting that our results are exactly in line with the view expressed in Landau & Lifshitz (1959) that "In general the constant (Bernoulli's constant) takes different values for different streamlines". Indeed we regard our calculations as confirming that nothing should be added to, or taken from, this statement. © European Southern Observatory (ESO) 1999 Online publication: December 4, 1998 |