The Hydrogen Clouds of M33

New Stromgren-beta Pleiades data on a zero point defined by the Hyades and Coma are presented. The new data and counterparts from four extant sources are compared to measurements published by Crawford & Perry in 1976. The only statistically significant problem found in this way is a 4 mmag offset obtained b - y from a data set published by McNamara in 1976. Results of similar comparisons published previously for Praesepe are updated by including three additional sets of published data. In this case, comparisons are made to measurements published by Crawford & Barnes in 1969. Two of the data sets (including one published by Joner & Taylor in 1995) yield accordant, statistically significant formal corrections to the b - y m(1) measurements of Crawford & Barnes. In addition, results from a third data set are consistent with those derived from the data of Joner & Taylor. However, this pattern is either partly or entirely absent from the offsets yielded by the other three data sets. Because the measurements of Joner & Taylor are still the only ones known to have been derived from direct comparisons among the Hyades, Coma, and Praesepe, it is concluded that the balance of evidence presently favors the offsets obtained from those data. In addition, it is noted that the data of Joner & Taylor do not imply that the data of Crawford & Barnes require corrections that depend on color. With these interim conclusions drawn, it is then suggested that further Praesepe measurements should be brought to bear on the problem.

A protocol for reddening analysis developed in previous papers in this series is reviewed here and applied to NGC 752. The protocol includes ( 1) detailed scrutiny of the histories of reddening determinations for program clusters, ( 2) adherence to statistical norms, ( 3) quality control of all reddening techniques, ( 4) a primary focus on techniques with limited metallicity sensitivity, ( 5) zero-point control of source data, ( 6) use of the solar neighborhood to establish a zero point for reddening values, ( 7) use of a standard permitting meaningful quantities to be as small as a few milli-magnitudes, and ( 8) rejection of ad hoc data assessments. After certain published results are set aside, it is found that a reddening value derived for F stars by using a technique that is insensitive to blanketing disagrees by about 0.04 mag with a value derived for K giants by using DDO photometry. This disagreement does not appear to be an artifact of the use of questionable data, and it persists if additional techniques are considered. Partly because two comparable reddening results for K giants are afterward found to disagree with each other, the problem is attributed to a blanketing effect ( with an unknown source) on the reddening values derived for the K giants. Pending a final resolution of that problem, the adopted value of E(B - V) for the cluster ( 44 +/- 3.4 mmag) is based solely on the initial F-star analysis. A discussion of the potential usefulness of small standard errors like the one derived here for E( B - V) is given. In a parallel investigation, it is found that no definitive value of [Fe/H] can be assigned to NGC 752 at present because a statistically significant disagreement between two published high-precision metallicities is detected.

From 1980 through 1998, the Wampler scanner and the 0.9 m Crossley reflector at Lick Observatory were used to obtain 137 nights of flux-curve measurements. In this paper, the resulting data are used to test, revise, and extend a set of flux curves for bright standard stars published in 1984. Full use is made of extensive measurements of a Lyr, the primary standard star. It is found that measurements made during spring and fall runs are highly consistent and that small systematic wavelength trends appearing in the standard-star data can be readily detected and corrected. This conclusion applies for 109 Vir in particular, and it is concluded that there is no convincing case for variation of this key standard star after 1979. Tests of the corrected and augmented flux curves using 13 color photometry yield a number of cases in which residuals are systematic with wavelength. However, for stars with E(B - V) < 0.12 mag ( including alpha Lyr), the revised and extended flux-curve data yield synthetic values of Stromgren u - b and b - y that are highly consistent with results from filter photometry. For almost all stars measured ( again including a Lyr), consistency with filter values of is ( V - R)(C) also obtained. Comparisons of the revised and augmented data with flux curves measured in the Soviet Union yield results ranging from fair agreement to marked disagreement. However, it is found that published data from additional sources include results that agree well with those presented here. Given this agreement, the results of the tests using filter photometry, and the character of the transformation coefficients for alpha Lyr, it is concluded that there is adequate support for the accuracy of the corrected standard-star flux curves. Recommendations about future measurements that can be used to test this conclusion are presented.

Since 1954, 70 reddening determinations have been published for M67. Those results are of very diverse quality, and they imply only that the value of E(B - V) for the cluster lies between -0.02 and +0.14 mag. In addition, this uncertainty is habitually concealed by long-standing citation practices. In response, a reformed reddening protocol called "alternative scholarship" is applied, with tenets including (1) rigorous published foundations for reddening values, (2) accountability for all extant reddening results for M67, (3) explicit quality control of reddening techniques, (4) use of techniques with limited metallicity sensitivity, (5) explicit zero-point control of input data (when possible), (6) a reddening zero point based ultimately on the dust-free solar neighborhood, and (7) adherence to statistical norms. Results from reddening maps are then set aside pending further examination of their zero points and accidental errors. In addition, reddening values derived from color-magnitude analysis and from averages of published reddening results are likewise set aside. Five techniques that satisfy the adopted protocol are then selected and applied. By averaging results from cluster A and F stars and K giants, one finds that the mean value of E(B - V) for M67 is 41 +/- 4 mmag. It is also found that extant results from additional techniques are consistent with this result or else can be set aside for plausible reasons. The analysis yields a collateral value of [Fe/H], and this is consistent with the derived reddening value. Using published high-dispersion data, it is found that the mean cluster value of [Fe/H] is -0:009 +/- 0: 009 dex. For M67 and the Hyades combined, no mean metallicity difference between giants and stars near the main sequence is found that is >= 0.036 dex at 95% confidence.

In this paper, transformations between Cousins R - I and other indices are considered. New transformations to Cousins V - R and Johnson V - K are derived, a published transformation involving T-1-T-2 on the Washington system is rederived, and the basis for a transformation involving b-y is considered. In addition, a statistically rigorous procedure for deriving such transformations is presented and discussed in detail. Highlights of the discussion include ( 1) the need for statistical analysis when least-squares relations are determined and interpreted, ( 2) the permitted forms and best forms for such relations, ( 3) the essential role played by accidental errors, ( 4) the decision process for selecting terms to appear in the relations, ( 5) the use of plots of residuals, ( 6) detection of influential data, ( 7) a protocol for assessing systematic effects from absorption features and other sources, ( 8) the reasons for avoiding extrapolation of the relations, ( 9) a protocol for ensuring uniformity in data used to determine the relations, and ( 10) the derivation and testing of the accidental errors of those data. To put the last of these subjects in perspective, it is shown that rms errors for VRI photometry have been as small as 6 mmag for more than three decades and that standard errors for quantities derived from such photometry can be as small as 1 mmag or less.

From histories of reddening studies for the Hyades, Praesepe, and nearby field stars, it is concluded that reddening scholarship requires fundamental reform. The adopted protocol for the reformed version of scholarship includes (1) rigorous published foundations for reddening values, (2) explicit quality control of reddening techniques, (3) use of techniques with limited metallicity sensitivity, (4) explicit zero-point control for both input data and derived reddening values, (5) an explicit choice of a reddening zero point, and (6) adherence to statistical norms. It is argued that only limited use should be made of reddening values from (1) spectral types for F0 and later type stars, (2) UBV analysis of blanketed stars, and (3) certain published relations between reddening and distance. However, reddening values may be based safely on (1) polarization measurements, (2) comparisons of beta and (R-I) C for F stars, and (3) Stromgren beta analysis of A stars. For the Hyades and Coma, upper 95% confidence limits on E(B-V) are found to be <= 1.0 and <= 3.2 mmag, respectively. For Praesepe, the derived value of E(B-V) is 27 +/- 4 mmag. The latter result is obtained after an allowance is made for the effects of an anomalous relation between b-y and (R-I) C in Praesepe A stars. A test of the possibility that differential stellar rotation causes the anomaly yields an inconclusive result. Adoption of the revised reddening value instead of the canonical value of zero yields an increase in the derived Praesepe distance modulus by 0.14 mag and a decrease in the derived cluster age by about 11%. In addition, by using published high-dispersion results, it is found that the value of [Fe/H] for Praesepe that is consistent with the derived reddening value is +0.01 +/- 0.04 dex. Further results for Praesepe stars will be required for a definitive comparison of the Hyades and Praesepe metallicities.