Positions, velocities, metallicities, and ages of stars are the raw material from which samples of the stellar populations in the Milky Way galaxy are defined. Correlations between these parameters provide clues to the chemical and dynamical evolution of our home galaxy. Yet, after decades of observation and study, controversy remains concerning the true relationships between these basic observational parameters even for the solar neighbourhood, and even more so concerning their interpretation. Thus, as models of the formation and evolution of the galactic disk improve, the need for more definitive observational data becomes ever more pressing.
Coordinated efforts are being made towards this goal on two fronts: On the
one hand, the chemical evolution of nearby solar-type stars has been studied in
detail by Edvardsson et al. (1993; E93) using high-resolution spectroscopy. In
parallel, a major programme has been undertaken to complete an inventory of the
solar neighbourhood in order to define large, homogeneous, complete, and
kinematically unbiased samples of nearby F and G dwarf stars (see, e.g.
Strömgren 1987; Nordström et al. 1996). Sample selection is
based on the
all-sky uvby-H
photometry by Olsen (1983, 1993, 1994a,b), which allows the
determination of distances, metallicities and individual ages. PPM proper
motions (Bastian & Röser 1991-1993) have been improved using modern
positions observed with the Carlsberg Automatic Meridian Circle
(Carlsberg
1991-1994). Finally, multiple radial-velocity observations have been obtained of
the F and G dwarfs in the sample, mostly with the CORAVEL scanners
(Mayor 1985).
Altogether, complete data have now been collected for some 10 000 stars in both
hemispheres. The programme is further described in Nordström et al. (1996).
Different subgroups of these stars are interesting for a variety of purposes. At the old, metal-poor end of the sample, the delineation and characterisation of the old thin disk, thick disk, and halo populations are the main interest. On the other hand, the early F dwarfs for which reliable ages and metallicities can be determined, but which are too young to have moved far from their place of formation, are of prime importance in clarifying the cause(s) of the scatter in the local age-metallicity diagram (see Sect. 5). In all cases, radial velocities are needed in order to compute space motions and galactic orbits for individual stars. In addition, repeated radial-velocity observations allow the detection of spectroscopic binaries. This is important because large-amplitude binaries might bias the kinematical parameters derived for the samples of stars to which they belong. In addition, unrecognised binaries with significant amounts of secondary light of a colour different from that of the primary may be classified incorrectly when photometric age or metallicity indicators are used to subdivide the initial sample. Detecting such binary systems and analysing their composite spectra is the first step in correcting for these effects.
The class of early F-type dwarfs contains a large proportion of rapidly
rotating stars (v sin i > 40 km s-1), for which radial velocities cannot be
accurately determined with CORAVEL. We have therefore observed that part of our
local sample with the Center for Astrophysics (CfA) Digital Speedometers
(Latham
1985, 1992), which can handle rotations up to 
120 km s-1
(Nordström et
al. 1994; Paper I), in order to determine their radial and rotational velocities
and search for binaries in the sample.
The present paper describes the selection of the sample and our observation and reduction procedures, gives results for 595 stars, including preliminary spectroscopic orbits for 18 double-lined and 2 single-lined binaries, and discusses the binary frequency in the sample and the extent to which such binaries may have been inadvertently included in the sample. Several forthcoming papers will further discuss the evolution of the galactic disk, based on the full data set (see, e.g. Nordström et al. 1997).