Comment on the dispersion-velocity of galactic dark matter particles

In a recent Letter, Cowsik, Ratnam and Bhattacharjee (1996a) have built a dynamically self-consistent spatial distribution of particles of galactic dark matter. They have come up with the rather unorthodox conclusion that the mean velocity dispersion of dark matter particles ``should be $600 kms$ or larger''. Their letter triggered immediate comments (Evans 1996, Gates et al. 1996). Here we find, as did Cowsik et al. (1996a), that models of the dark matter halo can be made consistent with velocity dispersion much larger than that expected from a simple application of the virial theorem in the solar neighbourhood. But in contrast to their conclusions, we show that using their model, we also obtain solutions with smaller velocity dispersion down to $\sim 270 kms$. These more orthodox dispersions arise because of less constraining boundary conditions for the central density but do not rely on indirect or model dependent measurements of the large scale distance behaviour of the rotation curve.