SUBDWARF STUDIES .2. ABUNDANCES AND KINEMATICS FROM MEDIUM RESOLUTION SPECTRA

The Ca II K line is observed in a sample of metal weak proper motion stars, and its strength is calibrated in terms of [Fe/H] from observations of previously well-studied stars. Radial velocities are also measured, and the three velocity components U, V, W are computed. We investigate the dependence of these velocities on abundance for stars with [Fe/H] < - 1.5. It is found that sigma-U, sigma-V, and (V) are independent of metallicity, but for the dispersion perpendicular to the Galactic plane we find sigma-w = 60 - 30([Fe/H] + 1). The last of these may be evidence for a vertical abundance gradient in the halo. We also find evidence in the published data of Carney et al. [AJ, 99, 201 (1990a)] for an abundance gradient in the perigalacticon domain. They noted themselves that their data do not show a metallicity gradient in apogalaction distance. Previous studies {e.g., Searle & Zinn [ApJ, 225, 357 (1978)]; Zinn [ApJ, 293, 424 (1985)]} have found no or very limited evidence for abundance gradients in the halo globular cluster system, although Seitzer [Ph.D. thesis, University of Virginia (1983)] did find a gradient, again in the perigalacticon domain. It may be, however, that the sigma-w vs [Fe/H] trend we see for field stars in a manifestation of the overlap of disk and halo populations around approximately - 1.6 less-than-or-equal-to [Fe/H] less-than-or-equal-to approximately - 0.6; this will be more thoroughly investigated elsewhere. We use a Monte Carlo procedure to compensate for the kinematic biases in the sample. Using both compensated and intrinsically unbiased estimators, we find the subdwarf system to lag the sun by 202 +/- 10 km s-1, which implies prograde rotation of the halo by 30 +/- 10 km s-1. We discuss models for the formation of the halo, and find our data to be easily reconcilable with merger scenarios. We also find an intriguing qualitative agreement with Yoshii's [A&A, 97, 280 (1981)] stellar surface accretion predictions. Our data do not support the existence of an exodus of subdwarfs from the solar neighborhood as has been suggested by Croswell et al. [AJ, 93, 1445 (1987)].