METALLICITIES FOR THE OUTER-HALO GLOBULAR-CLUSTERS PAL 3, 4, AND 14

Spectra at the Ca II triplet have been obtained of 11 giants in the outer-halo (galactocentric distances greater than 60 kpc) globular clusters Pal 3, Pal 4, and Pal 14, as well as of two giants in the more nearby cluster Pal 11. Similar spectra were also obtained of 38 giants in eight clusters with well-established abundances, allowing the calibration of a relation between Ca II triplet strength, V - V(HB), and [Fe/H]. This relation then yields abundances of [Fe/H]= - 1.57 +/- 0.19 for Pal 3, [Fe/H]= - 1.28 +/- 0.20 for Pal 4, and [Fe/H]= - 1.60 +/- 0.18 for Pal 14. Pal 11 is confirmed as a metal-rich cluster ([Fe/H]> -0.6), but a quantitative abundance could not be derived since the line strengths for this cluster lie outside the range of the calibration. The abundances for Pal 3 and Pal 14 are in good accord with previous determinations based on giant branch photometry, but that for Pal 4 lies between the two most reliable previous estimates, both based on giant branch photometry. Radial velocities were also determined for the giants in Pal 3, 4, and 14; our values agree with other recent determinations of the radial velocities of these clusters. Using our improved abundance determinations for Pal 3, Pal 4, and Pal 14, together with additional metallicities and distances from the literature, we reinvestigate the size of the abundance gradient present in the halo globular cluster system with either galactocentric distance (R) or distance from the Galactic plane (absolute value of Z). We find, at most, only shallow declines in mean [Fe/H] and no evidence at all for an abundance gradient for R > 8 kpc or absolute value of Z > 4 kpc out to the limit of the halo globular cluster system at distances of approximately 100 kpc. Even with our new abundances, the horizontal branches of these three outer-halo clusters are anomalously red for their metal abundances relative to inner-halo globulars, a manifestation of the well-known second parameter problem. If age is the second parameter, then we estimate that Pal 3 and Pal 14 must be approximately 1.5 Gyr and greater-than-or-equal-to 2 Gyr younger, respectively, than the representative inner-halo globular cluster M3. Similarly, Pal 4 must be at least as young as the famous second parameter cluster NGC 362. Color-magnitude diagrams that reach the main-sequence turnoff have been published for Pal 4 and Pal 14. We find that the position of the turnoff in Pal 4 is in accord with the age hypothesis for the second parameter and that the Pal 14 color-magnitude diagram is not inconsistent with it, although improved photometry is desirable in both cases.