CCD photometry on the intermediate-band vbyCaHβ system is presented for the metal-deficient open cluster NGC 2420. Restricting the data to probable single members of the cluster using the CMD and the photometric indices alone generates a sample of 106 stars at the cluster turnoff. The average E(b - y) = 0.035 ± 0.003 (s.e.m.) or E(B - V) = 0.050 ± 0.004 (s.e.m.), where the errors refer to internal errors alone. With this reddening, [Fe/H] is derived from both m1 and hk, using b-y and Hβ as the temperature index. The agreement among the four approaches is reasonable, leading to a final weighted average of [Fe/H] = -0.37 ± 0.05 (s.e.m.) for the cluster, on a scale where the Hyades has [Fe/H] = +0.12. When combined with the abundances from DDO photometry and from recalibrated low-resolution spectroscopy, the mean metallicity becomes [Fe/H] = -0.32 ± 0.03. It is also demonstrated that the average cluster abundances based on either DDO data or low-resolution spectroscopy are consistently reliable to 0.05 dex or better, contrary to published attempts to establish an open cluster metallicity scale using simplistic offset corrections among different surveys.

CCD photometry on the intermediate-band uvbyCaHβ system is presented for the metal-deficient open cluster NGC 2243. Restricting the data to probable single members of the cluster using the color-magnitude diagram (CMD) and the photometric indices alone generates a sample of 100 stars at the cluster turnoff. The average E(b - y) = 0.039 ± 0.003 (s.e.m.) or E(B - V) = 0.055 ± 0.004 (s.e.m.), where the errors refer to internal errors alone. With this reddening, [Fe/H] is derived from both m1 and hk, using b - y and Hβ as the temperature index. The agreement among the four approaches is excellent, leading to a final weighted average of [Fe/H] = –0.57 ± 0.03 (s.e.m.) for the cluster, on a scale where the Hyades has [Fe/H] = +0.12. Using a combination of photometric and spectroscopic data, 27 probable cluster members are identified and used to delineate the red giant branch and a well-defined clump at V = 13.70, while eliminating the so-called second clump at V = 14.1. Interpolation between isochrones of appropriate [Fe/H] leads to an apparent modulus of m - M = 13.15 ± 0.1 and an age of 3.8 ± 0.2 Gyr. A differential CMD comparison with Ber 29, a cluster with a galactocentric distance almost twice that of NGC 2243, constrains Ber 29 to be at least as young and as metal-rich as NGC 2243.

CCD photometry on the intermediate-band uvbyCaHβ system is presented for the open cluster, NGC 3680. Restricting the data to probable cluster members using the color-magnitude diagram (CMD) and the photometric indices alone defines a sample of 34 stars at the cluster turnoff that imply E(b-y) = 0.042 ± 0.002 (s.e.m.) or E(B-V) = 0.058 ± 0.003 (s.e.m.), where the errors refer to internal errors alone. With this reddening, [Fe/H] is derived from both m1 and hk using both b-y and Hβ as the temperature indices. The agreement among the four approaches is excellent, leading to final value of [Fe/H] = -0.14 ± 0.03 for the cluster and removing the apparent discrepancy between the past uvby analyses and extensive results from the red giants. The primary source of the photometric anomaly appears to be a zero-point offset in the original m1 indices. Using the homogenized and combined V, b-y data from a variety of studies transformed to B-V, the cluster CMD is compared with NGC 752, IC 4651, and the core–convective-overshoot isochrones of Girardi et al. By interpolation to the proper metallicity, it is found that the E(B-V), m - M, and age for NGC 752, IC 4651, and NGC 3680 are (0.03, 8.30, 1.55 Gyr), (0.10, 10.20, 1.7 Gyr), and (0.06, 10.20, 1.85 Gyr), respectively. The revised age and metallicity sequence and the color distribution of the giants provide evidence for the suggestion that the giants defining the apparent clump in NGC 3680 are predominantly first-ascent giants, as indicated by their Li abundance, while the clump stars in NGC 752, 0.1 mag bluer in B-V, are He-core–burning stars. When combined with the color distribution in IC 4651, it is suggested that over this modest age range where He-core flash becomes important, the distribution of so-called clump stars switches from being dominated by He-core–burning stars to first-ascent giants in the bump phase.

CCD photometry on the intermediate-band uvbyCaHβ system is presented for the old open cluster NGC 6253. Despite a high level of field star contamination because of its location toward the Galactic center, combination of the data from the multiple color indices with the core cluster sample derived from radial star counts leads to the identification of a set of highly probable, single cluster members. Photometric analysis of a select sample of 71 turnoff stars produces a reddening value of E(b-y) = 0.190 ± 0.002 (s.e.m.) or E(B-V) = 0.260 ± 0.003 (s.e.m.). The metallicity indices, δm1 and δhk, both identify this cluster as the most metal-rich object studied on either system to date. Simple extrapolation of the available metallicity calibrations leads to [Fe/H] values ranging from +0.7 to +0.9. Metal-rich isochrones with overshoot imply an age between 2.5 and 3.5 Gyr, with an apparent distance modulus between (m - M) = 11.6 and 12.2, depending upon the isochrones used. The improvement in the fit using α-enhanced isochrones may indicate that the cluster [Fe/H] is closer to +0.4, but the photometric indices are distorted by an elemental distribution other than a scaled solar. The Galactocentric position of the cluster, in conjunction with data for other clusters and Cepheids, is consistent with the inner disk reaching and maintaining a metallicity well above solar since the early history of the disk, unlike the solar neighborhood.

The effects of the orbital diffusion of stars and clusters on the Galactic abundance gradient have been modeled. For a single generation of field stars, the results confirm previous analyses based on the assumption of a linear gradient with Galactocentric distance. Though the dispersion in metallicity at a given radius grows with time, the underlying slope across the disk remains unchanged. For a disk dominated by a discontinuity in [Fe/H] beyond the solar circle, but no gradient within the inner or outer disk, a gradient is derived for any sample whose Galactocentric baseline overlaps the discontinuity. The narrower the range in distance around the discontinuity, the steeper the gradient. The consequence of orbital diffusion is to shift the observable edge of the transition in [Fe/H] toward smaller Galactocentric distance, enhancing the detection of a gradient. In contrast, star clusters of a single generation evaporate so quickly that the small surviving sample of the oldest clusters is statistically incapable of defining any gradient except a linear case. Inclusion of multiple generations of stars and clusters with a plausible age-metallicity relation enhances these patterns by expanding the metallicity range at a given Galactocentric distance through the inclusion of stars with a wide range in both age and Galactocentric origin. For field stars, the metallicity distribution of stars in the solar neighborhood is virtually indistinguishable for the two cases modeled. Because ages, abundances, and distances can be reliably determined for clusters and any sample of clusters is invariably dominated by those clusters formed over the last 2 Gyr, the cluster population offers a viable means for finding detailed structure within the recent Galactic abundance gradient. Using clusters to extend the Galactocentric baseline would supply an additional critical test of the discontinuity.

We have used an extended version of the Stromgren photometric system to study candidates in a search for extremely metal-poor stars.

The color-magnitude diagrams of the open clusters NGC 2420 and NGC 2506 have been investigated as intermediate links between the solar neighborhood and the Magellanic Clouds. Two sets of theoretical isochrones that include convective overshoot are zeroed to the Sun at solar abundance and to the unevolved main-sequence dwarfs of the Hipparcos Catalogue at [Fe/H] = -0.4, requiring a differential of 0.4 mag between the unevolved main sequences at a given color. Adopting E(B-V) = 0.04 and [Fe/H] = -0.39 for NGC 2506 and E(B-V) = 0.04 and [Fe/H] = -0.29 for NGC 2420, the respective apparent moduli are (m-M) = 12.70 and 12.15, while the ages of both clusters are approximately 1.9 +/- 0.2 Gyr or 2.2 +/- 0.2 Gyr, depending on the choice of isochrones. From the composite giant branch of the two clusters, the mean clump magnitudes in V and I are found to be +0.47 and -0.48 (-0.17, +0.14), respectively. Applying a metallicity correction to the M-I values, the cluster sample of Udalski leads to (m-M)(0) = 18.42 (+0.17, -0.15) and 18.91 (+0.18, -0.16) for the LMC and SMC, respectively. A caveat to this discussion and potentially to the claim that clusters of the same abundance and age are identical is the observation that the V-I colors of the red giants in NGC 2506 are significantly redder at a given B-V than the giants in clusters of comparable age and/or metallicity. The distance scale above has been derived using the general cluster relation between B-V and V-I. If the CCD photometry in NGC 2506 is correctly tied to the standard system, M-I for the clump will decrease and the distance moduli should increase by 0.1 mag.

New photometry on the uvby Ca system is presented for over 300 stars. When combined with previous data, the sample is used to calibrate the metallicity dependence of the hk index for cooler, evolved stars. The metallicity scale is based upon the standardized merger of spectroscopic abundances from 38 studies since 1983, providing an overlap of 122 evolved stars with the photometric catalog. The hk index produces reliable abundances for stars in the [Fe/H] range from -0.8 to -3.4, losing sensitivity among cooler stars due to saturation effects at higher [Fe/H], as expected.

The primary DDO metallicity indicators for stars in the [Fe/H] range of +0.3 to -1.1 have been recalibrated. It is found that the slope of the SCN metallicity transformation needs to be reduced while the zero-point is color-dependent: no color dependence is found for the slope. Over the [Fe/H] range from +0.3 Co -0.5, DDO precision photometry and delta CN provide [Fe/H] with an uncertainty of only +/-0.05 dex. Contrary to expectation, between [Fe/H]=-0.9 and -2.0, delta CN exhibits some sensitivity to changes in [Fe/H], though with a much steeper slope than at solar [Fe/H]. For the critical [Fe/H] range between -0.5 and -0.9, within the uncertainty the trend is consistent with little to no sensitivity to metallicity change. For C-0(3835), the isometallicity relations in the C-0(3845) - C-0(4548) diagram are nonlinear at the blue end. In contrast with delta CN, the metallicity sensitivity of C-0(3845) is high for [Fe/H] < - 0,4, but declines rapidly at higher metallicity. Using the delta CN calibration, the 16 nonvariable, red giant members in 47 Tuc exhibit a range in [m/H] from +0.2 to -0.7, consistent with a large spread in CN abundance, When divided into CN-strong and CN-weak samples, the mean abundances become -0.09+/-0.18 (s,d.) and -0.49 +/-0.11 (s.d.) for 6 and 10 stars, respectively, The mean abundance of the CN-weak stars is more than a factor of two higher than expected from the predicted [Fe/H](SPEC) of -0.9, in the C-0(3845) - C-0(4548) diagram, the 47 Tuc giants show no separation by CN strength, though the scatter among the bluer stars is large. Despite the scatter, [Fe/H]=-0.9 is clearly excluded by the data, while the mean of the points is best fit by the curve for [Fe/H]=-0.6+/-0.1. 47 Tuc appears to have a DDO metallicity which is a factor of two higher than found for field stars at the Same [Fe/H](SPEC); it is not representative of the field star population at the same [Fe/H]. Use of 47 Tuc to calibrate field star photometric parameters remains questionable, while identification of field star counterparts to 47 Tuc giants may not prove reliable. Possible explanations for the source of the discrepancy are discussed. (C) 1996 American Astronomical Society.

The spectroscopic abundances of Ryan & Lambert [AJ, 109, 2068 (1995)] are reanalyzed and shown to be consistent with the general and detailed conclusions of Twarog & Anthony-Twarog [AJ, 107, 1371 (1994)] regarding the DDO calibration and the metal-weak thick disk. The metal-rich red giants form a natural extension of the trend found among stars with [Fe/H](SPEC)<-1.1, without any evidence for a significant break near [Fe/H](SPEC)= - 1.2. Using an expanded sample comprised of only high dispersion spectroscopic abundances, it is demonstrated that the original DDO isometallicity relations based upon globular cluster giants differ in both slope and zero-point from the relations defined by the field giants. The globular cluster calibrators at intermediate metallicity exhibit systematically higher CN strength than the field stars at the same [Fe/H]. At the metal-rich end, the 47 Tuc isometallicity relation reveals a dope which differs from the field stars in the same manner as found for the intermediate-metallicity clusters. The hotter field stars with [Fe/H](SPEC) between -0.35 and -0.55 occupy the same region of the C-0(3845)- C-0(4548) diagram as 47 Tuc. The possibility of problems with the zero points of the cluster and field star photometry is also noted. (C) 1996 American Astronomical Society.

New photoelectric (U)BV(RI) data for 71 red giants and turnoff stars in the old open cluster Melotte 66 are presented and discussed. The photometry confirms the broad color range exhibited by giants at and above the level of the clump. UBV analysis of turnoff stars yields [Fe/H]=-0.53+/-0.08 assuming a cluster E(B-V)=0.16+/-0.02. A nearly identical [Fe/H] is implied by the giants' ultraviolet excesses, but with a significantly higher dispersion. Previously published CCD vby data have been transformed to the BV system to permit an analysis of the cluster color-magnitude diagram with reference to isochrones which incorporate convective core overshoot. Main-sequence fits imply an age of 4.5+/-0.5 Gyr and an apparent modulus (m-M)=13.75. Further support for these reddening and distance values is provided by comparison of the red-giant clump to other old open clusters and of the M(I),(V-I) data to a sample of globular and old open clusters. The giant branch of Mel 66 is shown to extend to an extreme in both luminosity and color currently matched only by old open clusters NGC 6791 and Cr 261.

A catalog of photometry of 1990 stars on the Caby system is presented. The sample includes stars covering an extensive range in spectral type, luminosity class, and metallicity, as well as apparent magnitude. The hk indices are on the standard system defined in Anthony-Twarog et al. (1991), while the V and (b -y) indices have been transformed to the system of Olsen (1993; hereafter referred to as 093), superseding the values in the original catalog of standard stars. Extensive comparisons to three earlier u V by catalogs with a large degree of overlap are made and the residual trends discussed. The qualitative sensitivity of the hk index as a function of temperature is illustrated for the primary spectral ranges.

Updated uvby observations for a larger sample of metal-deficient red giants arc presented and combined with a select sample of data from the literature transformed to a common system. Using the reddening maps of Burstein & Heiles [AJ, 87, 1165 (1982)], new absolute magnitudes, distances, metallicities, and reddenings are derived for each star. The metallicities are determined with a revised calibration of the m1, (b - y) diagram based upon comparison to a compilation of recent spectroscopic abundances transformed to a common system. The photometric abundances agree very well with the spectroscopic; the dispersion among the residuals for 58 giants is +/- 0.16 dex. The dereddened indices are used to show that for red giants with [Fe/H]< - 1.5, there is a well-defined relation in the c0, (b - y)0 diagram which exhibits only a weak dependence upon metallicity. Use of the standard relations allows one to obtain reddening estimates for normal halo field giants and globular clusters in the appropriate metallicity range.

DDO abundance estimates for metal-deficient field red giants from the survey by Norris et al. [ApJS, 58, 463 (1985)] are compared with high dispersion spectroscopic abundances and recalibrated uvby photometry. The DDO abundances are shown to deviate from the standard system by an amount which is metallicity dependent, i.e., a simple offset cannot transform the DDO data to the spectroscopic system. A recalibration of the DDO photometry demonstrates that the old DDO calibration provides reliable [Fe/H] estimates near -0.8 and -2.3, but systematically underestimates the metallicity of stars near -1.2 by about 0.5 dex. Reanalysis of the red giant sample of Norris et al. leads to a metallicity distribution much closer to that of the globular clusters while significantly decreasing the fraction of metal-weak disk stars. A possible explanation for origin of the discrepancy is that the clusters of intermediate [Fe/H] used in the DDO calibration are not representative of the field stars at the same [Fe/H].

A CCD study of the peculiar open cluster, Melotte 66, on the nu by Hbeta system is presented. The V,(b-y) color-magnitude diagram of this metal-deficient, old disk cluster confirms the anomalous features found in photographic BV studies and extends the main sequence to V=20. The main sequence exhibits a (b - y) color range significantly larger than expected from photometric errors alone; the (b - y) colors are correlated with Hbeta implying that the dispersion is real and is not the result of variable reddening across the cluster. The m1 indices for the turnoff stars imply a spread in [Fe/H] too small to explain the color range. Though the subgiant region is poorly defined, the region above the turnoff in the Hertzsprung gap is richly populated by a group of yellow stragglers whose membership is confirmed by radial star counts and radial velocities. The giant branch is richly populated but shows a smaller spread in color at a given V than one would expect from the turnoff region. More important, the giants exhibit a large scatter in m1, in contradiction with the turnoff observations. Various explanations for these observations are discussed, including CN variations among the giants triggered by either binary evolution or rapid rotation among the main-sequence stars.

New photometric and ultraviolet spectroscopic data for the metal-deficient red giants CD -38-degrees-245 and BD -18-degrees-5550 are presented and discussed in light of recently noted photometric anomalies. From the IUE spectra it is apparent that the ultraviolet excess found in BD -18-degrees-5550 is not the result of a hot companion. The IUE spectra, in conjunction with other observations, point to anomalous nitrogen abundances as the source of the discrepancies between the photometric and spectroscopic abundances for these stars. CD -38-degrees-245 appears to be exceptionally nitrogen-rich while BD -18-degrees-5550 is anomalously nitrogen-poor with respect to stars of comparable metallicities. While BD -18-degrees-5550 appears to be an exception to the rule for its metallicity, the confirmation of a similar photometric pattern for CS 22885-96 may be an indication that the nitrogen overabundance in CD -38-degrees-245 is typical for giants of extreme metal-deficiency. Additionally, the IUE spectra exhibit evidence for emission in Mg II in CD -38-degrees-245. This indicator of chromospheric activity could be linked to the long-term variability observed in the hk index centered on Ca II H and K.

CCD vby photometry has been obtained in six fields of NGC 6397 and used to delineate the principal sequences for this metal-poor globular cluster. Based upon the distribution in the m1, (b - y) diagram, a sample of highly probable members has been identified. From this sample, photometric binary candidates have been selected from the main sequence by comparison of their location in the color-magnitude diagram to the well-defined fiducial relation in V, (b - y). It is found that approximately 7% of the main-sequence members exhibit evidence for a composite nature. Though the photometric discrepancies appear real in that the fields are well away from the cluster core and the binary images do not suffer from any identifiable abnormalities, the binary fraction is consistent with an optical rather than a physical association of the superposed pair. For a reddening value of E(b - y) = 0. 14 and [Fe/H] = -2.0 an apparent distance modulus of (m - M) = 12.1 is derived if the parallax-calibrated main sequence of Laird et al. [AJ, 95, 1843 (1988)] is compared to the fiducial cluster main sequence, leading to an age greater than 18 X 10(9) yr from the luminosity of the turnoff. In contrast, a modulus of (m - M) = 12.4 is derived if the oxygen-enhanced isochrones detailed in McClure et al. [AJ, 93, 1144 (1987)] are adopted as the standard main-sequence relation; the corresponding age can range between 14 and 18 X 10(9) yr for only minor adjustments in the luminosity of the turnoff. Use of the differential approach of Sarajedini & King [AJ, 98, 1624 (1989)] places NGC 6397 in the old tail of a broad globular cluster age distribution.

We have obtained and analyzed CCD frames on the uvby system of a field in omega Centauri in order to constrain the intrinsic dispersion of heavy elements among its unevolved stars. Our results from direct measurements of the metallicity using m1 and from indirect measures using the color spread in (b - y) near the turnoff imply an intrinsic dispersion in [Fe/H] of about 1 dex, similar to the results of spectroscopic analysis of the evolved stars. The color spread near the turnoff is consistent with little spread in age, though the constraint is weak. The luminosity spread among the subgiants does allow a range of 2 x 109 yrs if the metal-rich stars are also younger. We have independently recovered the variable blue straggler E39 and have discovered an additional blue straggler candidate. Finally, we note that the variation in c1 as a function of magnitude for probable upper main sequence members suggests that previously evolutionary corrections for metal-poor stars may have been overestimated.

A fifth filter (fwhm = 90 angstrom) centered on Ca II H and K has been developed for use with the standard u-upsilon-by system. The filter, called Ca, is designed primarily for applications to metal-poor dwarfs and red giants, regions where the uvby metallicity index, m1, loses some sensitivity. An index, hk, is defined by replacing upsilon in m1 by Ca. The effects of interstellar extinction on the index are modeled and demonstrated to be modest and relatively insensitive to spectral type. Observations of V, (b - y), and hk for 163 primary standards are detailed and transformed to the standard V and (b - y) system. A qualitative analysis using only the primary standards indicates that hk is more sensitive than m1 over the regions of interest by about a factor of 3.