We have obtained high-resolution, high-signal-to-noise-ratio echelle CCD spectra of three red giants in the metal-poor globular cluster NGC 2298. A detailed analysis of their chemical composition yields [Fe/H] = - 1.91 +/- 0.1 (using a solar Fe abundance of 7.52), where the error includes the dispersion in the observed values and uncertainties in the model-atmosphere parameters. From spectrum synthesis of the [O I] line at 6300 angstrom, we find [O/Fe] = + 0.2 +/- 0.2, with all three giants having the same ratio within the errors. These stars fall somewhat below the mean trend between [O/Fe] and M(V) exhibited by the O-rich sample in Sneden et al. (1991). The alpha elements Mg, Si, and Ca are enhanced more than oxygen, with an average of [alpha/Fe] = - + 0.47, suggesting an oxygen depletion as might be expected from proton burning by the CNO tri-cycle. Al is strongly overabundant in all three stars relative to typical values found in field halo giants of similar metallicity. This Al overabundance could be due to proton burning of the envelope material by the Mg-Al cycle. The apparent dispersion in Al abundance would suggest that the Al overabundance is due to self-enrichment by the individual stars, rather than Al-rich primordial gas, but this conclusion requires verification. Based on our metallicity and composition we apply corrections to the estimated age of NGC 2298 found by previous studies, giving an age range of 14-18 Gyr. Our O and alpha-element abundances for NGC 2298, combined with those of other recent investigations of the most metal-poor globular clusters, indicate substantially lower values than measured by Abia and Rebolo (1989) or assumed O enhancements in the isochrones of McClure et al. (1987), leading to somewhat larger ages for these most ancient Galactic systems.

We report analysis of echelle spectra (R = 17,000, S/N almost-equal-to 50) of 12 Galactic bulge K giants in Baade's window from the sample of Rich ( 1988). We perform an abundance analysis for 11 stars ranging from [Fe/H] = -1 to [Fe/H] = 0.45. The accuracy of our abundance scale is confirmed relative to the disk super-metal-rich stars via the metal-rich giant mu Leonis, and to the Galactic globular clusters using a star in NGC 5927.

We analyze the neutron-capture element (Z > 30) abundance distribution of the ultra-metal-poor (but neutron-capture element rich) halo star CS 22892-052. The observed stellar elemental distribution is compared with those produced by the slow and rapid neutron capture processes (i.e., the s- and r-process) in solar system material. This comparison indicates that the elemental abundances, from barium to erbium, in this Galactic halo star, have the same relative proportions as the solar system r-process distribution. Within the uncertainties of the abundance determinations, the elements strontium and zirconium, but not yttrium, also fall on the same scaled solar r-process curve. The main component of the s-process cannot reproduce the observed neutron-capture abundances in this star. The weak component of the s-process, expected to occur during core helium burning in massive stars, can fit the relative abundance distribution of Sr and Y, but not Zr, suggesting that for the currently observed abundances in CS 22892-052, an admixture of the weak s- and the r-process might be required for production of the elements Sr to Zr. These results give evidence of the occurrence of heavy element nucleosynthesis, particularly the r-process, early in the history of the Galaxy, and further suggest a generation of massive stars (the astrophysical site for the r-process), preceding the formation of this very metal poor halo star, that was responsible for producing the observed heavy elements.

The metallicity of stars in the Galaxy ranges from [Fe/H] = -4 to +0.5 dex, and the solar iron abundance is epsilon(Fe) = 7.51 +/- 0.01 dex. The average values of [Fe/H] in the solar neighborhood, the halo, and Galactic bulge are -0.2, -1.6, and -0.2 dex respectively.

We analyze the recent thorium detection in the metal-poor halo star CS 22892-052. Sneden et al. have demonstrated that all of the stable elements with Z greater than or equal to 56, including those near the thorium nuclear region, are consistent with the solar r-process abundances. This result strongly suggests that thorium (formed in the r-process) was also produced in solar proportions in the progenitor of CS 22892-052. Theoretical calculations, presented here, that reproduce the observed stable solar system r-process abundances predict a thorium/europium (Th/Eu) ratio in close agreement with the extrapolated, corrected r-process-only ratio (0.463) at the time of the formation of the solar system, Sneden et al, found that Th/Eu = 0.219 in CS 22892-052, substantially below the current solar ratio, indicating a much greater age for this star. Ignoring additional production of thorium over time, and thus any chemical evolution effects, comparison between the observed and the solar system corrected Th/Eu ratios gives a simple radioactive-decay age for CS 22892-052 of 15.2 +/- 3.7 Gyr. Additional age estimates, based upon theoretically determined Th/Eu ratios, also suggest a range of 11.5 less than or similar to t(CS 22892-052) less than or similar to 18.8 Gyr and are thus consistent with that simple radioactive-decay age determination for this star. Since additional Galactic production of thorium leads to an increase in this decay time, this is a lower limit on the stellar age, and hence the age of the Galaxy. We evaluate the magnitude of this effect on the chronometric age determinations by employing several simple chemical evolution models, including a closed-box model and one which allows for Galactic infall. These Galactic chemical evolution models suggest an age of 17 +/- 4 Gyr for CS 22892-052. We discuss the implications and possible sources of errors and uncertainties in these age estimates.

New, improved, barium abundances for 33 extremely metal-poor halo stars from the 1995 sample of McWilliam et al. have been computed. The mean [Ba/Eu] ratio for stars with [Fe/H] less than or equal to -2.4 is -0.69 +/- 0.06 dex, consistent with pure r-process nucleosynthesis within the measurement uncertainties. Although the [Sr/Fe] and [Ba/Fe] abundance ratios span a range of 2.6 dex, the mean values are approximately constant with [Fe/H]. This is consistent with a model of chemical evolution in which the parent clouds were enriched by small numbers of supernova events. In this model, the decreasing heavy-element dispersion with increasing [Fe/H] is simply due to the averaging of element yields from many supernovae at higher [Fe/H]; however, it is necessary to increase the number of extremely metal-poor stars known in order to confirm this picture. In addition to the random Sr component from the r-process, the [Sr/Ba] ratios indicate that there is a second, also random, source of Sr from an as yet unidentified nucleosynthesis site.

We discuss how abundance ratios can help unravel the evolutionary history of the Galactic bulge. An LTE abundance analysis of a solar-[Fe/H] bulge red giant star reveals that the su-elements O, Mg, Si, Ca and Ti are enhanced by +0.3 dex, the r-process element Eu is enhanced by +0.5 dex and the Ba/Eu ratio is -0.5. These ratios are characteristic of normal halo composition, and suggest that the bulge reached solar [Fe/H] in less than 1Gyr.

In extremely metal-poor stars ([Fe/H]less than or equal to -2.5) the neutron capture elements are characterized by a 300-fold dispersion in M/Fe ratios which decreases with increasing metallicity, the median M/Fe ratio increases with increasing [Fe/H], but the average M/Fe number ratio is approximately constant. These observations are consistent with a highly dispersed intrinsic yield of neutron-capture elements in supernova (SN) events, and a progression to increasing metallicity by stochastic chemical evolution.