The southern sky gives us the great opportunity to observe two among the brightest and nearest globular clusters of the Galaxy: ω Cen and 47 Tuc. For these giant clusters, we present the comparison between observations and King-Michie multi-mass dynamical models with anisotropy in the velocity dispersion. A more comprehensive description of this work is to be published (Meylan 1986a,b).

Low mass X-ray sources observed in many globular clusters are usually interpreted as compact binaries with degenerate components (e.g., Hertz and Grindlay 1983). Degenerate stars can exist in globular clusters if the IMF contains a sufficiently large number of high mass stars. Since the main-sequence lifetime is a very steep function of stellar mass, most of degenerate stars can be regarded as primordial. If the typical mass of degenerate stars is higher than that of main-sequence stars, mass segregation makes the core crowded with degenerate stars. Tidally captured binaries between degenerates and main-sequence stars can abundantly form as the core density becomes very high.

We present calculations of globular cluster evolution performed by a modified Fokker-Planck approach, in which binaries formed by tidal capture are followed explicitly, along with subsequent heating mechanisms. The cluster is simulated by a two component model, using the cross sections of Press and Teukolsky (1977) for tidal capture, those of Hut (1984) for the single-binary encounters and for distant binary-binary encounters, and those of Mikkola (1983) for the strong binary-binary encounters. The initial state of the cluster is a Plummer model with N = 3 × 105 and scale radius ro = 1.13 pc. All stars are identical, with mass M∗ = 0.7M⊙ and R∗ = 0.57R⊙. This gives an initial core radius rc = 0.8 pc, and one-dimensional dispersion σ = 11.6 km s-1. All binaries are assumed to be identical, with separation a = 2.5R∗. There are no binaries in the cluster initially. Additional important effects, such as tidal truncation, tidal shocks, stellar evolution and mass loss, and stellar mergers, are not included.

Over the past decade, a very considerable amount of effort in stellar dynamics has gone into the study of interactions between binary systems and other stars. The asymptotic analytic results obtained by Heggie (1975) for binary-single star encounters have been largely confirmed and extended by later numerical experiments (Hills 1975, Hut and Bahcall 1983). Binary-binary interactions have been studied by Mikkola (1983).

We investigate the evolution of globular clusters in the Galactic tidal field prior to core collapse. These multimass models incorporate mass loss by stellar evolution.

Mathieu et al. (1986) have completed an extensive radial-velocity survey of over 100 late-type stars in M 67 with V < 12.8. The spatial distributions of the spectroscopic binaries and single stars (i.e. those stars without detected radial-velocity variation; many of these are undoubtedly binaries, albeit with lower secondary masses) are shown in Fig. 1. The distribution of the binaries is notably more centrally concentrated than the single stars. The two observed distributions derive from distinct parent distributions at the 98% confidence level. The projected half-mass radius of the binaries is 0.9 pc; the half-mass radius of the single stars is 2.4 pc. Indeed, 77% of the binaries lie within the single-star half-mass radius.

Forty-eight blue straggler stars have been discovered in NGC 5466, the only Galactic globular cluster known to contain an anomalous Cepheid of the sort found in dwarf galaxies. The stars were identified in color-magnitude diagrams constructed from photometry of deep photographic plates taken with the Canada-France-Hawaii 3.6 m telescope (calibrated with new UBV photoelectric photometry), and from point spread function photometry of CCD frames taken with the Palomar 5 m telescope. The stars typically have magnitudes <V> ~ 19.m1 and colors <B-V> ~ 0.m2. Forty-two of the 48 stars are situated inside of R=2.5 arcmin (see Fig.1), the projected radius containing half the cluster luminosity, and only six stars are found between 2.5 and 9 arcmin. A one-sided, two-sample Kolmogorov-Smirnov test (using the CCD data) establishes at the 98% significance level that the blue stragglers are more centrally concentrated than the subgiant stars of the same magnitude. By fitting multi-component King models to the projected radial distributions (Fig.2), the mean mass of the blue stragglers is shown to be ~1.5 to two times larger than the masses of the stars that contributed the light from which the core and tidal radii were derived (i.e. M (Blue Str.)=1.3±0.3 M⊙). Because the central relaxation time for NGC 5466 is much less than the cluster age, the different radial distributions are attributed to mass segregation. A similar mass segregation is also observed in the globular cluster NGC 5053, where Nemec and Cohen (1986, in preparation) have recently identified ~30 blue stragglers. The low stellar density and small escape velocity of NGC 5466 make a recent epoch of star formation (during which the blue stragglers might have formed as massive single stars) seem unlikely. Instead, the blue stragglers probably are either close binary systems that have transferred mass, or are coalesced stars. The very low frequency of stellar collisions expected in the center of NGC 5466 suggests that the blue stragglers are primordial binary systems. The simultaneous presence in NGC 5466 of the blue stragglers and the anomalous Cepheid V19, and their relative numbers, supports the hypothesis that there is an evolutionary connection between the two types of stars. By fitting theoretical isochrones to the photographic c-m diagram, NGC 5466 is estimated to have an age of 18±3 Gyr.

We report the identification of 23 faint blue horizontal branch stars in Omega Centauri similar to those discussed by Buonanno et al. (1985) in M15. We find that these stars are significantly concentrated towards the center of the cluster with respect to other giants. We suggest that they may have formed from the collision of a main sequence star and a white dwarf.

The four possible origins usually discussed for blue stragglers in stellar systems are (a) mass transfer in, or coalescence of, close binaries; (b) main sequence lifetime extension either through internal mixing or high non-thermal pressures; (c) complete mixing events in evolved stars at the core helium flash; and (d) ongoing star formation over long intervals. However, alternative (d) can be ruled out for the case of globular clusters because any residual gas in the cluster is very efficiently swept out as the cluster passes through the galactic disk, an event that occurs approximately every hundred million years. Hypotheses (a) and (b) both predict masses for the blue stragglers that exceed the turnoff mass but if alternative (c) is correct, then the blue stragglers should have masses less than or equal to the turnoff mass.

The identification of the optical counterpart of the M 15 (NGC 7078) X-ray source has been confirmed photometrically and spectro-scopically (Aurière et al. 1984, 1985, 1986; Ilovaisky and Chevalier: 1985; Charles et al. 1985, 1986). It is designated M15:AC211 from the Aurière and Cordoni catalogue (1981).

Two classes of globular cluster X-ray sources are known. Each consists of compact objects accreting material from a close binary companion. The brighter class has a neutron star primary, and the low luminosity class has a white dwarf primary. These sources formed by tidal capture of the compact object by a main sequence dwarf in the core of the globular cluster. Their presence and number has implications on the end points of stellar evolution in globular clusters and on the formation of binaries in cluster cores.

Two new methods for the detection of hot gas in globular clusters are proposed: 1. measurement of the position change of a radio source due to refraction and 2. determination of the concentration of hot gas using the Faraday rotation of the polarization plane of the background radio source radiation.

A 1-parameter family of anisotropic models is presented. They all satisfy the Plummer law in the mass density, but have different velocity dispersions. Moreover, the stars are not confined to a particular subset of the total accessible phase space. This family is mathematically simple enough to be explored analytically in detail. The family is rich enough though to allow for a 3-parameter generalization which illustrates that even when both the mass density and the velocity dispersion profiles are required to be the same, a degeneracy in the possible distribution functions persists. The observational consequences of the degeneracy can be studied by calculating the observable radial velocity line profiles obtained with different distribution functions. It turns out that line profiles are relatively sensitive to changes in the distribution function. They therefore can be considered to be more natural observables when a determination of the distribution function is desired.

It is often maintained that Antonov's equation, a linearization of the collisionless Liouville-Boltzmann equation, governs small perturbations of a stellar system. The variational integrals resulting from Antonov's equation are in six dimensional phase space. However, expanding the perturbations in the velocity coordinates and carrying out the integrals over the velocity components gives integrals in the three dimensional configuration space. Solutions in successive approximations lead to standing density waves. The first order equations involve a vector field ξ(x) related to but not identical with the Lagrangian displacements of a volume element of the system. In this respect the problem is analogous to the linear oscillations of a fluid star. The analogy is exploited to provide a classification for the modes of oscillation and to obtain suitable data for variational calculations. The normal modes appear to be trispectral, in the sense that the associated vector field ξ(x) is derived predominantly either from a scalar potential, a toroidal vector potential, or a poloidal vector potential. The eigenfrequencies of the radial (ι = 0) and non-radial (ι = 1) modes are calculated. The associated density waves are analyzed.

An isolated conducting gas sphere (an idealized model of a globular cluster) can have non-singular postcollapse similarity solutions for appropriate energy sources. Solutions have been found for a source mimicking three-body binaries; they are parametrized by N, the number of stars in the cluster, and are linearly stable if N < 7000, overstable if 7000 < N < 40,000, and unstable if N > 40,000. The nonlinear development is always oscillatory and produces a typically lower central concentration than in the “equilibrium”.

Recently, Fall and Rees (1985) have proposed a theory for the origin of globular clusters forming from the largely primordial gas in the protogalaxy. These authors have explained the typical masses of proto-globular cluster clouds (~ 106 M⊙) as gravitationally unstable condensations at temperature T ~ 104 in a hot protogalactic medium (T ~ 106 K) but they were not concerned with how these clouds would fragment into stellar masses (~ 1 M⊙). In fact, their proto-globular cluster clouds are trapped at T ~ 104 K, and cannot cool to lower temperatures. However, substantial cooling must occur if these clouds are to form solar mass stars. It is known that under primordial conditions the only available cooling agent is molecular hydrogen, formed in the gas phase. Therefore, if sufficient molecular hydrogen is formed, it is possible to cool the gas well below T ~ 104 K. In the following we outline how non-equilibrium conditions lead to a larger H2 abundance than derived by Fall and Rees, who assumed equilibrium conditions.

Behind the moving cosmic strings, wakes are formed from dark matter and baryon matter. The shock waves for baryon matter appear after the stage z≃790(μ/10-6). These wakes will fragment into lumps of mass M≃108(μ/10-6)3(vst/c)3/(1+z)1.5M⊙ where μ and vst~c are the line density in units of G=c=1 and the velocity of the strings.