We have determined elemental abundances for 47 G and K dwarf stars
with iron abundances ranging from solar to [Fe/H] =
0.4 dex. The stars
selected either move on relatively eccentric orbits spending most of
their time inside the solar circle, even having been born there, or
move on solar like orbits. Among the stars on solar like orbits there
are also a number of Extreme Population I stars, with photometric
logarithmic metallicities 
dex. These samples were
selected in an attempt to study differences in the chemical evolution
in different parts of the galactic disk. Our main results and
conclusions are summarized below.
The results fall roughly into two categories; galactic chemical evolution as studied from G dwarf stars and issues related to spectroscopy of dwarf stars and in particular to overionization in K dwarf stars.
Galactic chemical evolution
Our data do, in general, fit well into the overall picture of galactic chemical evolution as it is currently understood from empirical studies. They give further constraints on theoretical models of galactic chemical evolution for the region above solar metallicity, where metallicity dependent supernova yields further complicate the picture. In particular we find:
, nor any other element.
This indicates that the chemical evolution has lately been very
homogeneous in the regions under study. This is in contrast with
findings for 
-elements in the more metal-poor part of the disk
population, Edvardsson et al. (1993a),
Our lack of success in tracing variations in relative abundances,
e.g. of [O/Fe], [Na/Fe], [Si/Fe], [Ti/Fe] etc to vary with the
stellar velocities, i.e. with the present orbital mean distance from
the galactic centre (
). Thus, the probable mixture of the
metal-rich stellar sample, e.g., of one old and one more recent
population is not revealed as differences in relative
abundances. That may be because of several reasons. One is that the
relative number of stars from each population may not change very
strongly with galactic radius. Another may be that the yields of
heavy elements from the sites of nucleosyntheis are such that, e.g.,
the interstellar Na/Fe ratio may stay rather constant through the
evolution of the galactic disk. A third possibility is that galactic
orbital diffusion is strong enough to bring these differences below
the limit of detection. Thus, the observed scatter in such
abundance ratios can not be explained as the result of the mixing of
stars with different formation sites at different distances from the
galactic centre.
For oxygen and europium our data suggest a continuation in the declining trend found by earlier studies at lower metallicities. This suggests that enrichment from supernovae Ia is still important, also at these relatively high metallicities.
The scatter in titanium abundances found in Edvardsson et al. (1993a) is reproduced in our data in spite of the fact that we use more lines to determine abundances from. It seems probable, but remains to be proven, that this scatter is "cosmic''.
Stars with very high metallicities, so called Super Metal Rich, SMR, stars, have attracted some attention over the years. These stars have been claimed to either represent a late stage of the local galactic disk evolution or to be members of the Galactic bulge, which have now been scattered out to the solar neighborhood. Barbuy & Grenon (1990) obtained photometric metallicities and oxygen abundances for eleven such stars. They found these stars to have high photometric metallicities and to be over-abundant in oxygen. Castro et al. (1997) have obtained abundances for 9 similar stars. They find that oxygen, calcium and titanium relative to iron declines with increasing [Fe/H], while sodium, silicon and nickel relative to iron increases. Several of these stars have fairly low temperatures and we interpret their results for calcium and nickel as further evidence for the phenomenon possibly arising from overionization. No velocity data have so far been published for these stars and therefore it is difficult to compare with our results. We have, however, obtained spectra and derived abundances for three of the stars in Barbuy & Grenon (1990). These stars remain puzzling - their proposed very high [Fe/H] have not been verified, but at least some of them seem to have, as these authors suggest, a high [O/Fe] ratio. This should be further explored.
Overionization
For the K dwarf stars we do detect what appears to be a case of overionization for, in particular, iron and chromium, such that abundances derived from FeII and CrII are much higher that those derived from FeI and CrI. FeI abundances do not vary with the effective temperature. Also for calcium and nickel the K dwarf stars depart markedly from the trend outlined by the G dwarf stars. Signs of similar departures for rare earth atoms are also traced.
Acknowledgements
Erik Hein Olsen is thanked for providing his catalogue prior to publication, invaluable for the planning of the project and subsequent analysis of the data. Poul-Erik Nissen is thanked for valuable comments and suggestions and for taking time to teach SF how to most efficiently reduce echelle spectra. Pierre Magain is thanked for compiling and sharing the individual line-to-line scatter for the studies from his and Zhao's work. Bengt Edvardsson is thanked for providing programs for the photometric calibration and comments during the work.
Karin Eriksson carried out a substantial fraction of all the measurements of equivalent widths and Matthias Palmer and Mikael Nilsson the determinations of surface gravities from the strong 6162 Å CaI line. We are much indebted to all three for these contributions.
The Swedish Royal Academy of Sciences is gratefully thanked for giving SF a grant to travel to the U.S.A. and thus being able to partake in the observational work at McDonald Observatory. This work was carried out while BG held the Beatrice Tinsley visiting professorship at the University of Texas, which is also gratefully acknowledged. The Observing Committee at McDonald Observatory generously granted observing time.