The HR diagram is a useful shorthand locating a star in a two-coordinate space. For the astrophysicist, the y-coordinate is bolometric luminosity, Mbol, the x-coordinate, effective temperature, Teff. Objects of given chemical composition, age (or evolutionary status) are labeled in the xy plane by mass. For an observer, y may be apparent or absolute magnitude in a certain wavelength region and x may be spectral type or color. The HR diagrams for populations differ because of age, chemical composition and stellar masses present. HR diagrams are often of mixed nature; some involve observables others derived or semi-theoretical quantities. I will display various types of HR diagrams for low-luminosity stars. For galactic or extragalactic studies the HR diagram needs a further dimension, the frequency of stars at an x,y. The mass of the Galaxy, but not its light, may be dominated by M dwarfs. HR diagrams are also interesting for their nearly empty spaces. In Fig. 1 we show as a sample, the basic results of the U.S. Naval Observatory parallax program, in which broad band (B-V) colors define the visual luminosity, My, on the main (MS) and degenerate (WD) sequences.

The status of determinations of white dwarf radii by model atmosphere methods is reviewed in this paper. Details will appear elsewhere (Shipman 1978). In brief, the results are that (i) the mean radius of a sample of 95 hydrogen-rich stars with parallaxes is 0.0131 R⊙; (ii) the mean radius of a sample of 13 helium-rich stars is 0.011 R⊙, indistinguishably different from the radius of the hydrogen-rich stars; and (iii) that the most serious limitation on our knowledge of the mean radius of white dwarfs is the influence of selection effects. An estimate of the selection effects indicates that the true mean white dwarf radius is near 0.011 R⊙.

The lone dot in the lower left of Russell's original diagram (1914) was disregarded, since it did not fit into his suggestive scheme of giant-to-dwarf evolution. Bottlinger (1923) - in his nearly forgotten study of photoelectric color indices - seems to be the first who, with the evidence of the then three classical white dwarfs, invoked a new equation of state and clearly rejected all other possibilities of explanation. Kuiper (1941) summarized the knowledge obtained during the thirties, mainly by his spectroscopic work. A decade later Luyten (1952) presented impressive HR diagrams based on admittedly crude color indices and, for many more objects, on reduced proper motions.

The group of subdwarf O stars consisting of field stars and some central stars of old planetary nebulae does occupy an intresting place in the HR diagram. Greenstein and Sargent (1974) have tried to establish this place, and conclude that especially the hottest ones need ultraviolet data to improve the values of effective temperature and absolute luminosity.

Eggen (1969) has defined a subluminous star “… as one that is fainter by at least a factor of 10 in visual luminosity than the main sequence stars of the same(U-V).” Implicit is the suggestion that single stars might exist in the region of the HR diagram between the subdwarfs and the normal degenerate sequence. While the evolutionary calculations of Chin and Stothers (1971) indicate that stars with masses <0.2 M⊙ will pass through this region following the completion of nuclear burning, these models have not yet been confirmed. We summarize Naval Observatory astrometry and photometry as it pertains to the existence of such stars.

The determination of the distribution of white dwarfs in the HR diagram is marred by two observational difficulties: The recent discoveries of degenerate stars with unusual spectral characteristics and the controversy over the existence of crystallization sequences. Observationally, the origin of the controversy can be traced to the single band displayed on the MVvs. (b-y) diagram by several white dwarfs with accurate parallax (Graham 1972). The discrepancy between this single band and the two bands evident in the MVvs. (U-V) diagram (Eggen and Greenstein 1965) is far too great to be attributed to observational errors (Eggen 1969). Theoretically, the number of sequences in the HR diagram is directly related to the cooling properties of degenerate dwarfs, for which crystallization is expected on rather general grounds. On one hand, Van Horn (1968) considered a discontinuous transition to the solid phase with the release of latent heat and the addition of it to the total internal energy. Since the amount of latent heat depends upon the chemical composition, bright and faint crystallization sequences should be observed according to the different stages of evolution reached by the parent stars. On the other hand, Mestel and Ruderman (1967) proposed a gradual transition to the solid phase, and consequently the formation of a single cooling sequence. The purpose of this contribution is to report observational results indicating the existence of two crystallization sequences of DA type white dwarfs.

When analyzing the spectral energy distributions of early type stars as measured at Cerro Calán, we find that many of them have a feature which we have called ΔD (Gutiérrez-Moreno et al., 1968); the internal error of the measure of ΔD for a single star is ±0.02. An analysis of the spectrum scans and the energy distributions shows that this feature may be interpreted as an emission in the Balmer continuum. A few stars have values of ΔD<0, and a comparison of their scans with those of normal stars (ΔD=0) suggests an absorption larger than normal in the Balmer continuum.

Photoelectric and photographic UBVRI photometery have been obtained for over 700 stars in the field of Ara OB1, including many in the central cluster NGC 6193. Association members were identified mainly on the basis of their reddenings. The association sequence is well populated blueward of (B-V)O = −0m.10, (U-B)O = −0m.35. Beyond this “turn-up” point stars lie only above the ZAMS. There are clear gaps in the color magnitude diagrams just redward of the turn-up points. The width of these gaps is 0.07 mag. in (U-B). They cannot be easily explained by pre-main sequence models involving only gravitational contraction or by those involving 3He burning. The difficulty of identifying the faint pre-main sequence stars (if they exist) amidst the field stars is discussed and a limiting luminosity function for the association is derived.

Intermediate band photometric data for stars in at least 50 open clusters now exist. This paper reviews the status and desiderata for the uvbyß subset. Similarities and differences between clusters are highlighted, and the relation of the photometric differences to physical differences is noted. Some of these are stellar age, chemical abundance, rotational velocity, and duplicity.

It is fitting to mark the Centennial of Henry Norris Russell's birth with a discussion of a new approach to the calibration of absolute magnitudes for MK spectral types. With the addition of the NGC 6231 main sequence down to A0, the material for the cluster fitting method using very carefully determined MK types is complete. The data will be published elsewhere (Garrison and Schild 1978).

Theoretical zero age main sequences (TZAMS) from Flower (1976) are compared with two observed zero age main sequences (ZAMS) in Fig. 1.: ZAMS40, based on a distance of 40 pc (Sandage 1957) and ZAMS46 on a distance of 46 pc (see Flower 1977 for Teff: (B-V): BC scales). Although ZAMS46 lies above any TZAMS, the TZAMS can be shifted in luminosity by changing the chemical composition.

We have classified on the MK system a total of 455 stars in 12 open clusters and associations. The classification is based on wide (1.2 mm) spectra of two reciprocal dispersions (39, 128 Å mm−1) obtained with the Kitt Peak 2.1 m and 90 cm reflectors, respectively. The higher dispersion is necessary to show the subtle peculiarities found in some stars. The clusters are the Orion Nebula cluster, Orion OB1 association, Lacerta OB1 association, IC 2602, IC 4665, Pleiades, M39, M34, NGC 2516, NGC 6633, NGC 6475, and Coma; two of these were done with Morgan.

The clump giants in old and middle-aged open clusters provide a sample of stars whose evolutionary state can be clearly identified as core helium burning (Cannon 1970). The evolutionary state of the Ba II stars, on the other hand, is at present undetermined. The enrichment of s-process elements in the atmospheres of these peculiar stars suggests that the stars have at least passed through double shell burning, where the heavy elements may be produced and mixed to the surface. The Ba II stars are, however, too faint to be associated with this phase of stellar evolution; the absolute magnitudes and temperatures are consistent with the core helium burning phase. The barium stars should then occur in the clumps of Population I clusters in similar proportion to their numbers among the field giants.

Photometry of the K line of calcium in A-type stars is useful in segregation of Am stars and in estimating the metal abundance of field stars and of stars in clusters. To date, K-line photometry data for six clusters have been analyzed using uvbyß data: Hyades, Pleiades, IC 2391, and IC 2602 (Hesser and Henry 1971), Orion (Hesser, McClintock, and Henry 1977), and Praesepe (Henry, Anderson, and Hesser 1977). We present similar analyses for the Coma and NGC 6475 star clusters, using the uvbyß data of Crawford and Barnes (1969) and of Snowden (1976) respectively. The Coma stars closely resemble field stars. NGC 6475, on the other hand, is most remarkable. Of 31 stars for which complete data are available, 9 are metal-rich, 3 are metal-poor, and 19 (about 2/3 of the stars!) are Am.

The old disk clusters represent a striking contrast to the globular clusters which will be discussed extensively at this meeting in that the disk clusters show few anomalies. This makes a discussion of their abundance measurements of particular interest at this meeting, and may shed some light on the cause of the peculiar behavior of the halo objects. It is also appropriate to discuss the ages of the clusters since the HR diagram is a major tool for obtaining the ages of stars. A complete sequence of stars of one age in a cluster can be fitted to the ZAMS of theoretical models, and accurate abundances averaged over numerous stars in the cluster can be estimated. We will discuss the ages and abundances of 11 disk clusters for which UBV and DDO photometry are available in the literature. This limits the sample to those for which good color-magnitude (C-M) diagrams exist. In addition we present the C-M diagrams of two clusters that we have studied and that are not yet published elsewhere. We will use these clusters to illustrate our procedures.

Evidence for non-homogeneity in the metal abundances of giant stars in various globular clusters is reviewed. The nonhomogeneity is of two kinds: (1) differences in the value of from star to star in the same cluster, for particular values of A, and (2) non-constancy of the cluster mean value of as a function of A. Observational evidence for the existence of variations of both kinds is cited, and discussed for four “nucleosynthetic” groups of elements: viz., s-process, Fe-peak, α-process, and CNO. The rival hypotheses of core-envelope mixing and primordial abundance variations are examined in the light of the observations.

Until very recently, estimates of the chemical composition at early epochs of the Galaxy were made through the analysis of halo field stars. Recently, however some medium dispersion studies of stars in globular clusters have appeared. It would be expected that both approaches would lead to identical results. However, a quick glance at the literature gives the opposite impression.

During the past decade it has been evident that the morphologies of globular cluster color-magnitude diagrams cannot be described by only two parameters, age and metal abundance (Sandage and Wildey 1967, Hartwick 1968). The third parameter is unlikely to be helium because a normal helium abundance is needed to explain the pulsation of RR Lyrae stars (Deupree 1977) and because the helium abundance of the planetary nebula in the metal-poor cluster M15 is normal (O'Dell et al. 1964).

Photometric metal abundances of individual red giants in eight extremely distant halo globular clusters and the Draco and Ursa Minor dwarf spheroidal galaxies have been obtained using the Washington broad-band system, C, M, T1, T2 (Canterna 1976). Observations were made at the KPNO 2.1-m and CTIO 1.5-m telescopes. In Table I we list for each system the mean value of [Fe/H], the number of stars observed in each system, n, the Galactocentric distance, RGC, the intrinsic color of the giant branch at the level of the horizontal branch (HB), (B-V)o,g, and the fraction of HB stars bluer than the RR Lyrae gap, fB. Sources for unpublished color-magnitude diagram (CMD) data are: Pal 11 (Canterna and Schommer), Pal 12 (Canterna and Harris), and Ursa Minor (Schommer, Olszewski and Kunkel).

Most CM diagrams of well observed galactic globular clusters have giant branch widths which are understandable in terms of realistic estimates of random (and systematic) error associated with data acquisition and reduction. Not Omega Centauri. The scatter in (B-V) color among Omega Centauri giant branch stars is considerably greater than that allowed by observational error, differential reddening, or star-to-star mass and/or helium abundance variation (Demarque and Geisler 1963; Woolley 1966; Dickens and Woolley 1967; Rood 1973; Cannon and Stobie 1973; Cannon and Kontizas 1974; Iben 1974; Norris and Bessell 1975; Hesser et al. 1976; 1977). This problem is an old one, but one which has recently received a flurry of attention. Indeed, our next speakers have investigated possible causal relationships between interior mixing and Omega Centauri giant (or subgiant) color. I see from the program that they intend to discuss such things today, so we shall try to avoid those topics as much as possible.