Modern kinematic diagnostics from high-S/N spectroscopy provide useful tests on the collisionless merger hypothesis for the formation of elliptical galaxies. Bender, Saglia & Gerhard (1994) show that the skewness of the line-of-sight velocity distributions (LOSVD) has opposite sign to the rotation velocity and follows a distinct pattern. Previous N-body experiments (Bendo & Barnes 2000, Naab & Burkert 2001) disagree as to whether the relation is reproduced by the models.

We present first results from an ongoing effort to explain the skewness properties of the LOSVD's of ellipticals. We use large N-body simulations of disk-disk galaxy mergers ($10^6$ particles), with mass ratios 1:1 and 3:1. Our 1:1 models fail to reproduce the H3 vs. $V/\sigma$ relation, in agreement with Naab & Burkert (2001). 3:1 mergers yield trends similar to the observations for large values of $V/\sigma$, but with overall lower H3 values in the models than in real galaxies. And we fail to reproduce high H3 values for low $V/\sigma$. So far, it is unclear whether the disagreement is related to the lack of rotation in the bulge components of our initial model galaxies, or to a more general shortcoming of dissipationless mergers to reproduce the properties of real ellipticals.To search for other articles by the author(s) go to: http://adsabs.harvard.edu/abstract_service.html