Tubbs (1980) and later Simonson (1982) have produced a model of NGC 5128 where the dust lane represents material captured from the exterior and the are comprised of material not yet settled either in the principal plane of a prolate elliptical, or in the plane defined by the minor and intermediate axis of a triaxial one. In this way the warps are transient phenomena and they are destined to disappear and give rise to an elliptical with a straight dust lane, if tumbling is not present. An alternative interpretation has been given by van Albada, Kotanyi and Schwarzschild (1982), who were able to reproduce the phenomenon as a stationary one by assuming a slow tumbling motion around the minor axis of the triaxial figure of the elliptical body. However, the model requires that the tumbling motion has to be retrograde with the respect to the motions of the warps.
Four minor axis dust lane ellipticals with warps have so far studied, viz. Anon 0151-498, NGC 5128, and NGC 5266 (where the observed motions along the major axis are prograde), and NGC 5363 (with retrograde motion) (Bertola et al. 1985). If tumbling is present, the streaming motions could be either in the same or in opposite way (Freeman 1966, Vietri 1986) so it appears difficult to discriminate between the transient and stationary case. While a satisfactory model has been proposed for NGC 5128 on the first hypothesis, an equally satisfactory alternative model for NGC 5266 assumes counterstreaming with a velocity at 7 kpc twenty times larger than that of tumbling (Varnas et al 1986). As a consequence NGC 5266 would be an almost oblate (10% of triaxiality) galaxy with stable warps. Stability is also suggested by the ordered structure of the dust ring in this galaxy. Of course this does not mean that the gas and dust were not acquired. It is just a matter of how much time has elapsed since the encounter.
There are some other factors that could help in deciding the mechanisms responsible for the warps, as in the case of NGC 5363. In this galaxy the dust is so highly warped in the outer regions that the dust lane runs parallel to the major axis. To explain this in terms of transient warps would require the galaxy to be almost prolate. On the other hand the high rotational velocity suggests rather an almost oblate object. This contradiction is not present in the stationary model.
It is a pleasure to thank the Anglo-Australian Observatory for the unique opportunity which has been given to the author in obtaining the high resolution photographs of Fig.1 through their service photography. In preparing the manuscript I benefited discussions with M. Capaccioli, G. Galletta, M. Vietri and W. Zeilinger. Thanks are due to Peter Usher for reading the manuscript.