A classical astronomical problem has been to derive the total density of matter local to the Sun from observations of the velocity and density distributions of well-defined “tracer groups” of stars near the Galactic poles. Since the original work by Oort (1932: ρ0=0.15 M⊙/pc.3), the dynamical value of ρ0, has consistently been in excess of the amount of matter actually observed ; a substantial part of the mass would seem to exist in a form which renders it undetectable.
Our earlier survey of the dynamics of NGP A and F stars yielded 0.14 M⊙/pc.3 for ρ0 , (Hill et al. 1979) but employed an analytical potential derived by Camm (1950,52) requiring a (physically implausible) increase in the velocity dispersion with z-distance, although no such departure from the isothermal state was apparent in the velocity data for the F stars to z=200 pc. Such problems could well arise from the numerous simplifying assumptions involved in solving the combined Poisson and Boltzmann equations analytically. Recent studies (eg. Bahcall 1984) have used more sophisticated multicomponent mass models and solved numerically for the potential, taking as input the observed velocity dispersions and scale parameters for the various components.