The Laws Observatory of the University of Missouri is an example of the small university observatory that under lean financial circumstances succeeded for a number of years (1890-1920) as a successful research institution. Productive research appears to have been due to the temperament of the individual astronomers who had the ability to find projects which could be accomplished with small telescopes, minimal equipment, and few other resources. In these circumstances, a handful of students began careers in astronomy at the University of Missouri.

In five papers written between 1951 and 1955 Shapley considers the topic of red giant stars and reddish variable stars in the Magellanic Clouds. These works coincide with Shapley's final year as Director at Harvard and the first years of his retirement which extended a full score of years before his death in 1972. They include the following: Magellanic Clouds II (Supergiants/Red Variable Stars in the Small Cloud; January 1951); Magellanic Clouds IV (On Period Frequency Anomalies; February 1952); Magellanic Clouds VII (Star Colors and Luminosities in Five Constellations; March 1953); Magellanic Clouds VIII (On the Populations Characteristics of the Two Clouds; October 1953); and Magellanic Clouds XVI (Infrared Stars and Stellar Evolution; July 1955). These five papers, which appeared originally in the Proceedings of the National Academy of Science, may be found in the Harvard Reprint Series I as numbers 346, 360, 373, 376, and 425.

A new grid of evolutionary sequences has been computed for the main sequence and first red-giant branch (RGB) phases of low to intermediate mass stars. From these sequences we have obtained new intermediate age isochrones.

We report new results from a program which is aimed at obtaining deep CCD photometry for a sample of relatively nearby globular clusters having a wide range of metallicities. The CCD cameras on the CFHT 3.6 m, CTIO 4 m and KPNO 4 m telescopes have been used over the past 4 years to obtain deep exposures in regions of a number of clusters. In order to avoid the severest crowding, all of our observations have been obtained at distances of greater than ~ 5 core radii from the cluster centers. The images have been analysed by using the DAOPHOT point-spread-function fitting routines.

Multimass dynamical models are used to study the difference between a mass function determined in an outer field and the global mass function of the cluster. The differences are small.

We have made an analysis of the visual photometric data contained in the Catalogue of Concentric Aperture UBVRI Photoelectric Photometry of Globular Clusters (Peterson 1986). Structural parameters have been obtained by use of the Simplex algorithm of Caceci and Cacheris (1984) to fit the model curves of King (1966) to the run of cluster luminosity with radius. We find that concentric aperture photometry alone can be used to determine globular cluster core radii and central surface brigtnesses reliably. Application of this techique, however, is limited to about two-thirds of the known clusters of the Galaxy because no or inadequate numbers of photometric measurements exist for the remaining clusters. Accurate determination of cluster concentration classes still requires use of other types of data, such as star counts.

The non-spherical appearance of globular clusters was first noted by Pease and Shapley (1917) and discussed in some detail by Shapley (1930) who analyzed cluster shapes determined from star counts made by Helen Sawyer on the Franklin-Adams star charts. This classic work has provided most of the data set used in all subsequent discussions of cluster shapes. A number of studies reporting cluster shapes have appeared in the years since we began this project, the most recent of which include Geyer, Hopp and Nelles (1983), Frenk and Fall (1982), and Kadla et al. (1976, 1977).

We have completed abundance analyses of stars in three globular clusters: M71, M4. and M22. Spectra of resolution 0.3 and 0.6 (two pixel) resolution have been obtained with the Palomar coudé spectrograph and a TI CCD. The analysis was carried out with model atmospheres and f-values derived from three sources: absolute f-values derived by theory for the 6300 line of OI and for CN bands, laboratory f-values for lines that are too weak in the sun to be useful, and solar f-values. The last introduce an uncertainty of about 0.25 dex because solar f-values derived via the Holweger-Muller model differ from those derived via the BEGN model.

We think that it is possible to find the correct scale of abundance for metal-rich globular clusters thanks to the new generation of spectrographs, equipped with CCD cameras. We analyzed giants in ten globular clusters and Arcturus using high dispersion spectra acquired through the CASPEC spectrograph at the 3.6 m telescope at La Silla. The detector was an RCA CCD. Stars cooler than 4150 K were avoided since their absorption spectrum is too strong. By a comparison with standard Arcturus spectra, we found a small trend to overestimate equivalent widths. This systematic error affects the derived abundances only marginally. However, too large equivalent widths must produce too large metal abundances. Abundances were derived following a standard procedure.

In 1979 a disturbing controversy arose in the field of globular cluster research when Cohen (1980) and Pilachowski, Canterna, and Wallerstein (1980) announced the results of the first high dispersion studies of the composition of giants in the globular clusters M 71 and 47 Tucanae. In contrast to earlier studies, which found metallicities of typically −0.3 and −0.5 dex, these investigators obtained values of −1.3 and −1.1. Since then, many have attempted to redetermine the abundances of M 71 and 47 Tuc to explain the discrepant results. These efforts have all suffered from the absence of high signal-to-noise, high resolution spectra of stars with temperatures above 4300 K.

Integrated photographic image-tube spectra have been obtained for the central regions of a number of metal-rich Galactic globular clusters with a wavelength resolution of 2.5 å over the wavelength region λλ3400-4500 å. The spectra have been analyzed using a variety of quantitative spectral indices that compare the strengths of neighboring absorption features. Our main result is that two parameters are needed to describe the integrated spectra of metal-rich globular clusters. The second parameter is manifested in two ways:

(1) In a diagnostic diagram sensitive to surface-gravity the metal-rich clusters do not form a well-defined linear sequence. Instead, we find large differences from one cluster to another in the mean surface gravity of the stars contributing to the integrated light at 4000 å. It appears that the relative amounts of light contributed by dwarfs and giants varies considerably from one cluster to another.

(2) In a diagram that discriminates CN strength we find large differences from one cluster to another in the mean CN strengths of clusters having similar “spectral type”. It is inferred that the mean CN anomaly varies from one cluster to another and that the anomalous CN strengths are present in main sequence stars as well as in giants.

The above two second parameter effects are shown to be strongly correlated in the sense that the “dwarf-dominated” clusters have the largest CN anomalies. Hence, our study indicates that no more than two independent parameters are required to explain the integrated spectra of metal-rich globular clusters.

Details of the above findings can be found in Rose and Tripicco (1986, A.J. 92, 1610).

As a result of their investigation of 27 galactic globular clusters with the ANS satellite van Albada, de Boer and Dickens (1981) classified M 79 (NGC 1904) as an “extremely blue” cluster. It was also found to be a low luminosity x-ray source based on data acquired with the Einstein X-ray Observatory (Grindlay 1981). In this brief paper we discuss the stellar spectra extracted from two short wavelength (SWP) IUE images acquired at the “center of light” of M 79. Discrete peaks in the “spatially resolved”, cross-dispersion profile suggested the presence of at least three hot stars in the large aperture in both images (Fig. 1). The images in question are SWP 25303 and SWP 28936 for which it is important to note here that 1) the orientation of the large aperture on the plane of the sky differs for the two images by nearly 180° and 2) the target coordinates, determined from offset maneuvers with respect to a nearby SAO star, are nearly coincident. The cross-dispersion profile observed in the first spectrum is thus repeated in the second case, but is “flipped” left for right. Features immediately apparent in both images include a broad, asymmetric peak, almost certainly the blend of two or more components, and a second peak well separated from this blend and most likely a single star. In Fig. 2 we show the cross-dispersion profile of SWP 28936 and the best fit to that profile, assuming three components. From fits such as this and similar ones in other wavelength bins, we ultimately obtain from each image the entire SWP spectrum of each individual component.

The globular cluster ω Centauri is known to be chemically inhomogeneous, a property that reveals itself via a wide giant branch in the V, B-V color-magnitude diagram (Cannon and Stobie 1973). Typically a color range of δ(B-V) = 0.3 − 0.4 exists among giants with V < 13. However, in the red R, R-I (Norris and Bessell 1975, Bessell and Norris 1976), and I, V-I (Lloyd Evans 1977) diagrams a much tighter giant branch is seen. This has lead to the suggestion (e.g. Bessell and Norris 1976) that molecular absorption in the bandpass of the B filter has significantly reddened the B-V color, and thereby produced an anomalously wide giant branch.

New synthetic horizontal branch (HB) models are presented for some globular clusters known to have bimodal HB distributions. These models are based on new Yale HB evolutionary tracks for Y=0.25 and the core masses appropriate for the compositions. The distribution of stellar masses along the HB is given by a slightly modified version of Rood's (1973) function. Figure 1 compares the synthetic and the observed color-magnitude diagrams and the generalized histograms of the distribution of HB stars over (B-V)o (observational data from Alcaino & Liller 1984, Buonanno et al. 1981, Stetson 1981, and Menzies 1974 for clusters M4, M5, N1851, and N6723 respectively). Following Norris (1981), we have used the period-luminosity-color relations to estimate the colors of some RR Lyrae variables. Since none of the parameters in the models has a bimodal distribution, the excellent agreement between the color distributions of the models and the observations suggests that the observed bimodal distributions are a consequence of the evolution from the zero-age HB. Contrary to the suggestion of Norris (1981) and Smith & Norris (1983), there is no need to connect the bimodality of the HB with the observed bimodal CN distributions of the red giants in some of these clusters. The clusters in Figure 1 span a narrow range in metallicity ([Fe/H]=-1.40 to −1.09; Zinn 1985), and we are investigating other, less well observed, clusters in this range (e.g., N2808) to see if their bimodal HB distributions also have simple explanations.

We present preliminary versions of log g - log Teff diagrams for a number of clusters with gaps in their blue horizontal branches and for two normal clusters.

We discuss observations of He lines at λ 4026 å in two stars in NGC 6752 and three in M 92. M 92 and NGC 6752 are moderate to extreme second parameter clusters. The presence of He lines in stars where diffusion should deplete atmospheric helium may be a clue to the identity of the second parameter. One such candidate which could easily inhibit diffusion is rotation.

Spectra, at a dispersion of ~ 50 å per millimeter, have been obtained of BHB stars in the globular clusters M 3, M 13 and M 92 with the TV scanner on the Soviet Union's Six Meter Telescope. The spectra cover a range of 700 ångstroms in 500 channels in which counts were made of the intensity of the stellar spectrum. At this dispersion the hydrogen Balmer lines (γ, δ, ε, H8 - H12) can be seen as well as the Ca II line at λ = 3934.

Strömgren four-color photometric measures have been made of blue horizontal-branch A stars in the globular clusters M 4, M 13 and M 55 with the Steward Observatory 90 inch telescope and with the 60 inch telescope at Cerro Tololo Inter-American Observatory. These stars are faint, ranging in V magnitude from 13.6 in M 4 to 15.5 in M 13 and the corresponding errors in the four-color indices are ± 0.04 to 0.06 in the c1 index, for one observation. The error of the mean value of the c1 indices is approximately ± 0.02 for most of the stars since they have been measured from 4 to 10 times each.

The Vilnius photometric system was developed for photometric two-dimensional classification of stars and for the determination of interstellar reddening. The system consists of seven magnitudes U, P, X, Y, Z, V and S with mean wavelengths of 345, 375, 405, 466, 516, 544 and 655 nm and half-widths of the order 20 – 30 nm (Straižys 1977). Later on the system was successfully used for the determination of temperatures and the metallicities of halo stars (Bartkevičius and Sperauskas 1983).

Metallicities have been determined for a chemically unbiased sample of field halo dwarf stars. Their metallicity distribution function is similar to the predictions of a simple model of chemical evolution, but somewhat different from that of globular clusters.