Abundance ratios in a common proper motion pair: Chemical evidence of accreted substructure in the halo field?

Elemental abundances are presented for the metal-poor ([Fe/H]=-1.50) common proper motion pair HD 134439 and HD 134440. The abundances for the two stars are in very good agreement, with the neutral species showing only a small difference(similar to 0.05 dex) which is well within the statistical and T-eff uncertainties. The essentially identical abundances, kinematics, and parallaxes of the two stars indicate that they share a common history. This history, however, appears to be different than other metal-poor stars. Suggestions, based on kinematic evidence, that these two-stars are representative of a distinct accretion event are corroborated by our abundance ratios, which indicate [Mg/Fe], [Si/Fe], and [Ca/Fe] are consistently some similar to 0.3 dex lower than the vast majority of metal-poor field stars. Such underabundances have been predicted in environments like dwarf Spheroidals and the Magellanic Clouds. Moreover, our abundance ratio deficiencies are consistent with those recently observed in the the anomalously young globular clusters Rup 106 and Pal 12, which have been alleged to have been accreted from the Magellanic Clouds. The [Fe/H] and retrograde motion of the common proper motion pair are characteristic of the subset of Galactic globular clusters suggested by Rodgers & Paltoglou [ApJ, 283, L5 (1984)] to have been coalesced from satellite galaxies. We also call attention to the metal-poor subgiant BD+03 740 as another possible representative of an accreted or chaotically formed member of the halo field. If recent Fe analyses of this star are correct, then [Mg/Fe] and [O/Fe] are 0.5 dex lower than in other metal-poor field stars. This star also has a relatively low photometrically inferred age; relative youth has been noted as a possible characteristic of accreted field populations, and is qualitatively consistent with the young ages of the purportedly accreted globular clusters Rup 106, Pal 12, Ter 7, and Arp 2. Additionally, the revised [O/Fe] ratio for BD+03 740 would suggest a large spread, perhaps 0.7 dex, in [O/Fe] of field stars of very low [Fe/H]; this itself might provide strong evidence of some degree of chaotic halo formation in independent fragments. If, on the other hand, earlier Fe analyses of this star are correct, [Mg/Fe] and [O/Fe] for this star are unremarkable; however, the low gravity estimates from earlier studies would then suggest that BD+03 740 is a less than or equal to 3 Gyr field star with [Fe/H]similar to-3. Further spectroscopic study of this interesting object is needed to determine if it may be similar to the metal-poor ([Fe/H]=-3.1) high velocity star CS 22873-139, which Preston [AJ 108, 2267 (1994)] has argued is less than or similar to 8 Gyr in age. Finally, our abundance ratios for HD 134439 and HD 134440 suggest that low [alpha/Fe] may be a characteristic of accreted halo systems including the anomalously young globulars. However, as has been noted by others, the lo tv alpha-element abundances apparently cannot explain differences between photometric and Ca II-based metallicity estimates for these clusters, nor the variation in these differences between Rup 106 and Pal 12. (C) 1997 American Astronomical Society.