COMPARISON OF N-BODY SIMULATIONS TO STATISTICAL OBSERVATIONS OF GALAXY PAIRS

N-body simulations were conducted of pairs of galaxies with a 3:1 mass ratio on parabolic orbits in order to quantify the effect of dynamica l friction. The effects of varying the ratio of the dark matter halo s ize to the distance of closest approach were explored. Once the dark m atter halos are fully overlapping the more massive simulated galaxies achieve a larger maximum separation after the first encounter, despite the increased dynamical friction caused by the more extended halos. P rojected separation and radial velocity histograms were generated by ' 'observing'' the simulation results at various times and from various orientations. These histograms were compared with observations of gala xy pairs (Charlton and Salpeter 1991; Chengalur, Salpeter, and Tertian 1993) with the result that large halo radii (similar to 200-600 kpc) and wide distances of closest approach are generally favored. It is di fficult to reconcile the small radial velocity differences that have b een observed (median of similar to 30 km s(-1); Chengalur et al. 1993) with the simulations when we sample all parts of the orbits equally. Including an additional population of wide pairs that have just recent ly reached ''turnabout'' from the Hubble flow would lower the median v elocity differences. Models suggest that additional data for pairs at intermediate separations should have a somewhat larger median velocity difference than the wide pairs. Very narrow pairs include galaxies th at are interacting and whose gaseous components respond to forces othe r than gravity. If consistently small Delta v are measured from neutra l hydrogen velocities in a larger sample of narrow pairs, pressure for ces and dissipation effects on the gaseous components could be respons ible.