Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Close Binary Stars in Globular Clusters BRUCE MAROON University of Washington ABSTRACT Although close binary stars are thought theoretically to play a major role in globular cluster dynamics, virtually no non-degenerate close binaries are known in clusters. We review He status of observations in this area, and report on two new programs which are finally yielding candidate systems suitable for further study. One of the objects, a close eclipsing system in Cen, is also a blue straggler, thus finally providing firm evidence that globular cluster blue stragglers really are binary stars. INTRODUCTION A growing body of theoretical work indicates that a small number of close binary stars dominate the dynamic evolution of globular star clusters fag, Elson et at 1987~. For convenience, we here define a "close" binary as one with an orbital penod of a few days or less. Such systems may have binary orbital velocities of several hundred lon s~i, compared with a typical single star caster orbital velocity of just a few In sol. Therefore, literally just a handful of close binaries in a globular cluster can store as much kinetic energy of motion as possessed by the total of all 105 single stars in the cluster! This energy may be liberated by three-body encounters of cluster single stars with the binaries, and this source term is likely to be the dominant counterbalance of ldnetic energy sinks of stars escaping from the cluster. Cluster binaries may be pnmordial, or formed later through a variety of encounter processes, especially during the collapse of the cluster core due to the so called "gravothermal catastrophe". 270
HIGH-ENERGY ASTROPHYSICS 271 Elson et al. (1987) have pouted out the analogy of this energy source to thermonuclear fusion in stars: in this case, the cluster collapses, forms binaries, and then "burns" the orbital kinetic energy via encounters. A problem that observers may have with this elegant theoretical scenario is that virtually no close binary stars are known in globular clusters! Only one spectroscopic or eclipsing close binary, an object reported later in this paper, is known in any globular cluster. A small fraction of globular clusters are known to contain neutron star binary systems, evidencing themselves as either the X-ray bursters near the cluster centers, or as millisecond radio pulsars (although note that several of the latter systems are known to have binary periods of months rather than days, and thus store negligible kinetic energy). As only the very smallest fraction of globular cluster stars end their lives as neutron stars, however, one might expect binaries containing white dwarfs to be far more abundant in clusters (Verbunt and Meylan 1988), and systems with giant, subgiant, and main sequence components yet more common. SEARCHES FOR EVOLVED GLOBULAR CLUSTER BINARIES White dwarf close binaries in globular clusters are probably most easily found if they are undergoing mass exchange. The analog of such a system in the field, outside of clusters, is the cataclysmic variable (CV). Of course, both novae and dwarf novae easily draw attention to themselves during their intense optical outbursts, but even in quiescence such systems can be located through a variety of techniques. The optical spectra of CVs in quiescence are highly distinctive (strong, broad emission lines), as are the colors (distinct ultraviolet excess). A decade of study has also revealed quiescent CVs to be prominent soft X-ray sources, with luminosities in the range 103° through 1032 erg s~: (C6rdova and Mason 1983, 1984~. Novae are relatively rare in the field, and so one might expect only a very small number in globular clusters. Indeed, only two such objects have ever been reported: one in NGC 6093 in the year 1860 (Pogson 1860; Luther 1860), and one in 1938 in M 14 (Hog" and Wehlau 1964; Shara et al. 1986), and cluster membership for the latter, although probable, is still uncertain. However, dwarf novae are more common in the field (Patterson 1984), and one might hope that the analog of these mass exchange red dwarf~white dwarf systems might be found in globular clusters. In fact, only two dwarf novae are known in globular clusters. These stars, both discovered long ago during outbursts, have been recently recovered in quiescence: V101 in M5 (Margon et al. 1981), and V4 in M30 (Margon and Downes 1983~. These two objects are poorly studied to date; although at least V101 is known to stat be undergoing outbursts (Share et al. 1987), neither has ever been caught in outburst by a spectroscopist, and neither has been observed
272 AMERICAN AND SOVIET PERSPECTIVES for X-ray emission. Thus definite examples of white dwarf close binaries in globular clusters remain rare and poorly understood. The suggestion by Hertz and Grindlay (1983a, b) that the low- luminosibr X-ray sources observed outside of the cores of several globular clusters by the Einstein Observatory are mass-exchange close binary systems containing a white dwarf, analogous to the field cataclysmic variables (CVs), is therefore of great interest. These authors derive a luminosity function which implies that the galactic globular cluster system contains a total of ~103 such objects, a conclusion also reached by Droll (1984) and Hertz and Wood (1985~. EXOSAT observations have confirmed a few of these sources (Verbunt et al. 1986; Koch-Miramond and Auriere 1987; Auriere et al. 1989~. It is sometimes not amply enough stressed that all of the evidence which identifies these systems as white dwarf close binaries is quite indirect, based primarily on the fact that these sources have considerably lower X- ray luminosities than the highly luminous sources located within a few arcseconds of the cluster cores; the latter are now widely agreed to be neutron star binary systems (Lewin 1980; Lewin and Joss 1983; Grindlay et al. 1984~. In particular, in spite of several sensitive attempts, there has never been an optical identification of a single one of these low-luminosity cluster X-ray sources. Is there reason to suspect this evidence, and/or do reasonably plausible competing models exist? Verb unt et al. (1984) have pointed out the curious fact that all of these "low luminosity" sources, If in fact cluster members, have X-ray luminosities an order of magnitude larger than the classical U Gem systems in the held, despite their supposed physical similarity to those systems. Furthermore, if these systems share the most basic properly of the known dwarf novae (be., optical outbursts) it seems surprising that there could be a population of 103 of them in the Galaxy, and not one has ever been reported in optical outburst, despite a century of intensive photographic study of globular clusters for variable stars. Crowding is not likely to totally suppress discovery of the outbursts, if the sources reported by Hertz and Grindlay (1983b) are indicative of the class; the majority of them are many core radii from the cluster centers, and inspection of catalogs of cluster variable stars All, Hogg 1973) show that there are numerous single variables cataloged equally close to the core, at magnitudes equivalent to that that would be reached by a globular cluster CV in outburst. The two known cluster CVs cited above were in fact initially found exactly this way, despite their location in relatively crowded regions of the cluster. Verbunt et al. (1984) have suggested that the sources may instead be quiescent X-ray transients. Some of these sources may have an even simpler explanation: Hertz and Gnndlay (~1983a) calculate that ~3.3 of the 8 sources they discuss may be background QSOlAGN superpositions
HIGH-ENERGY ~TROP~ICS 273 or foreground stellar coronal sources. Furthermore, three clusters are reported to have spatially extended X-ray emission as well (Hartwick et al. 1982), raising the possibility that some of the reported point sources are confused or nonexistent. Margon and Bolte (1987) showed that if the X-ray sources reported by Hertz and Grindlay (1983a, by in the cluster Centaun were indeed like the field dwarf novae, they might be optically identified due to their peculiar colors; yet these workers were unable to locate any such counterparts (see also Shara et al. l988~. We do know that the cluster CV M5 V101, with (U-B) =-0.9 in quiescence,: does show the distinctive color of the field CVs, so this negative result is significant. Thus, the nature of the low-luminosity cluster X-ray sources remains terribly uncertain, and it is far from obvious that all, or even many, are truly close binary stars. Cluster binaries with red giant components at least have the virtue of being bright enough for easy spectroscopic observation. Gunn and Griffin (l979), working at a radial velocity precision of 1 An s~i, failed to find even a single binary candidate amongst extensive observations of 111 giants in M 3 spanning five years. However, a few successes are finally emerging. Pryor et al. (1988a) have shown that vZ 164 in M 3 is probably a binary with a period of a few years, and Pryor et al. (1989) report a handful of further candidate binaries, all with periods of years. Although these systems are at least true cluster binaries in the classical sense, their very long periods of course imply that they store negligible orbital kinetic energy, and cannot contribute to the energy budget of cluster, the problem which introduces this paper. The only other candidate for a short period globular cluster binary of which we are aware is V78 ~ the direction of ~ Cen (Bailey 1902~. This eclipsing Algol with a 1.2& period appears frequently in the literature with comments of varying degrees of certitude that it is a cluster member fag, Geyer 1967, 1971; Sistero 1968), but based on radial velocity it is now known definitely not to be a cluster member (Geyer and Vogt 19783. The very limited success in finding cluster binaries with evolved com- panions, either degenerate stars or red giants, has led my colleagues and me to initiate a number of programs aimed at finding binaries with main sequence componenm. We believe we have met with success through two different approaches. AN ECLIPSING BLUE STRAGGLER IN A GLOBULAR CLUSTER Blue stragglers, especially in globular clusters, have long rankled both 1 Multicolor photometry for MS V101 appears in Richer and Fahlman (1987), although the ob- ject is anonymous in that paper; it is the most extreme ultraviolet point in their Figure 11.
274 AMERICAN AND SOVIET PERSPECTIVES observers and theorists; their position on the color-magnitude diagram, with normal main sequence color and luminosity, but lying far hotter than the cluster tumors, is simply inconsistent with stellar evolution. Explanations abound, but there are few hard facts to back up these theories. Specifically, for more than 20 years they have been said to be close binaries ~IcCrea 1964; Hoyle 1964), with their odd colorlluminosity resulting from an ex- tensive past episode of mass transfer, but there is no direct evidence for this: not a single globular cluster straggler is an eclipsing or spectroscopic binary. Alternative theones, involving long-lived single stars undergoing odd evolution, are also viable (Wheeler 1979~. Some evidence that blue stragglers in NGC 5466 and NGC 5053 have a peculiarly concentrated spatial distnbution, and thus a higher primordial mass than other evolved stars in the cluster, has been provided recently (Nemec and Harris 1987; Nemec and Cohen 1989~. As has been previously stressed, however, this does not directly discriminate between binary mass exchange and mixing theories (Mathieu and Latham 1988~. Some hope of addressing the problem arose when the eclipsing system NJ~ 5 was identified In the field of co Cen, and it was pointed out that the star lies near the blue straggler area of the cluster color-magn~tude diagram (Niss et al. 1978~. This object may be identified with entry no. 5876 in the Woolley catalogue (Woolley 19663, and star 5642 of Dickens et al. (1988~. The period is 1.4&, and there is a deep (1.2 may) primary eclipse. However, a strange twist of fate appears to have quenched interest in the system: based on early photometry and the shape of the light curve, it was concluded that NJL 5 is probably not a cluster member (tiller 1978~. No spectra have ever been published for this object, although one unpublished spectrum has been cited as compatible with membership (Jensen and Jorgensen 1985~. As it was obtained in the green and shows only two spectral lines (Jorgensen 1987), it must surely be regarded as inconclusive. We obtained spectra of NJL 5 on five nights in 1988 at moderate `2Ay spectral resolution, using the RGO spectrograph and IPCS detectors at the Anglo-Australian Telescope. The bright cluster horizontal branch star ROA 4153 (Cannon and Stobie 1973), of color similar to NJL 5, was used as a radial velocity standard. The average of the five velocities of Nob 5 at various phases throughout the 1.4& photometric period is 8 ~ 7 hen s~i relative to the comparison star velocity, and thus, given the high velocity of Cen, 232 lan s-i (Meylan and Mayor 1986), the object is most certainly a cluster member. Our individual absolute velocity determinations for both NJL 5 and ROA 4153, although less precise than the cross correlation results, also clearly indicate that both stars are cluster members. NJL 5 is also almost certainly a blue straggler: the spectral type is near AS, and it is far below the cluster horizontal branch (DaCosta et al. 1986~. The spectrum is quite ordinary: the dominant features are Balmer adsorptions,
HIGH-ENERGY ASTROPHYSICS 275 with both stellar and interstellar Ca K lines also visible. There is no sign of a secondary spectrum. We have also obtained one low resolution red (~5500 - lo,000A) spectrum of NJL 5 with the Faint Object Red Spectrograph at the AAT, and again see no obvious sign of a secondary, although considerably more extensive phase coverage is desirable to strengthen this conclusion. As a historical footnote, we point out that the (B - V) color for the star listed in the Woolley (1966) catalogue differs from the normal color of the system by 1 may, for reasons which are unclear. A correct color appears in more recent photometry (Jensen and Jorgensen 1985; Dickens et al. 1988). NJL 5 now becomes very important: not only is it the first and only globular cluster blue straggler Mown to be a binary, but the very large eclipse depth shows the inclination is near 90°, and so the orbit is soluble and all system parameters may ultimately be obtained. Our existing moderate resolution spectroscopy is already tantalizing. Our five spectra have good phase coverage, and no radial velocity variations are convincingly detected; a conservative upper limit on the orbit K velocity is 30 km sol. This already implies a very high mass ratio and an exotic, undermassive secondary (recall that the unseen companion has radius comparable to the AS star to cause the deep primary eclipse). Thus it appears that the solution of the NJL 5 orbit may confirm the widely held but unproven scenario that blue stragglers result from a previous episode of very extensive mass transfer. There are several bright Algols in the field that have periods and light curves similar to NJL 5 and presage the probable parameters: RY Aqr (Heft 1987) and RT Per (Mancuso et al. 1977) both have q ~ 0.2, with K~ 30 and 60 km s-l, respectively. The low amplitude we infer for the radial velocity variations is presumably responsible for the failure of many previous observers to detect the binarity of blue stragglers in old open clusters (Stryker and Hrivnak 1984~. We have already obtained and are currently analyzing the necessary spectroscopy at much higher resolution to define the orbital parameters of NJL 5. This work on NJL 5 was done collaboratively with Russell Cannon, and will be described more fully elsewhere (Margon and Cannon 1989~. FAINT MAIN SEQUENCE BINARIES IN GLOBULAR CLUSTERS What are the prospects for detection of main sequence binaries of later spectral type than blue stragglers, and thus at a less odd stage of evolution? A conceptually simple way of identifying such binaries in globular clusters has been known for many years (Steinlin 1956; Fernie and Rosenberg 1961~: if a cluster color-magnitude (C-M) diagram can be obtained with very high photometric precision, and if nature cooperates with a metallicity spread
276 AMERICAN AND SOVIET PERSPECTIVES
HIGH-ENERGY ASTROPHYSICS 277 on the cluster ages, metall~cities, chemical homogeneity, etc., have been published for NGC 362 (Bolte 1987a) and NGC 7099 (Bolte 1987b, 1989~. The photometry, obtained with the CIIO 4m telescope, reaches far down onto the main sequence, and we do indeed find that its intrinsic width is unresolved at our level of precision, thereby enabling a search for the "parallel" main sequence due to putative binaries. For each of our four clusters, we have fields at three different core radii, potentially allowing a search for mass segregation as well. The quality of these data very substantially exceeds anything in the previous literature on these clusters, and a second main sequence separated by ~0.75 mag in luminosity would clearly stand out, greatly exceeding both the uncertainties in our photometry and the intrinsic (as yet unresolved) main sequence breadth. Sadly, no such sequence is apparent in any of the fields in any of our four clusters, with one exception: the most central field in NGC 288. Perhaps with hindsight, NGC 288 is an ideal globular cluster for a search for main sequence binanes. It has a modest distance modulus, ~14.5 may, permitting precise photometry far down on the main sequence. There is virtually no foreground contamination at this galactic latitude (b = - 89°!~. Most important of all, the object, although unquestionably a globular cluster by any definition, is sufficiently sparse to be virtually transparent almost all the way into the core. Thus, crowding is minimal in the innermost field, exacter where one expects the binaries to sink In Figure 1 we show our C-M diagram of the central field in NGC 288. One sees immediately approximated one dozen stars displaced upward in luminosity from the otherwise cleanly defined main sequence. The displaced stars are spatially well separated by our reduction programs, so we have no particular reason to suspect that their distinguishing characteristic is merely poor quality photometry. As an interpretive aid, we show with the solid line where the fiducial main sequence would fall if displaced upward in luminosity by 0.75 may, as expected for a sequence of equal mass binaries. The dozen outlying objects fall quite precisely on this sequence. This coincidence may be a conspiracy of nature, but if so it is particulars cruel that it occurs in the one field where we should have the best chance of seeing the binary sequence, nameh,r the most central field we possess with precision photometry. We thus do not suggest that NGC 288 is in any way unique, but merely that if all clusters in fact possess a few of these objects, it is quite logical that we would see them in this cluster first. One would be naive to believe that the objects offset from the main sequence in Figure 1 are binaries strictly on the grounds of our photometry. Surely either eclipses or radial velocity variations must be required to con- vincingly argue that the first population of globular cluster main-sequence binaries has been discovered. Eclipses would be obsenationally straight- forward to search for, but statistically unlikelier to be observed. A negative
278 V AMERICAN AND SOVIET PERSPECTIVES 16 18 20 22 24 I I I I I I r . I I I 1 o o _ o o o o o O O o O O O ~ O ~` o ooo oo o o o o o O ~' 00 0 0 O ~ O 0 0 O _ | l l I L o o o o 1 1 1 NGC 288 o o ~ o - _ ~r ~ ~ °~8 oo~co o ° o °o~O. o~OO 0 0 o8°~°~°O - o o° _ o o o - .5 (B V) 1 FIGURE 1 A color-magnitude diagram for NGC 288, denved from CCD photomet~y obtained at the prime focus of the 4m telescope of the CeIro Tololo Interamerican Obsenratory. I-he solid line is a locus off~;et by 0.75 mag ~om a fiducial main sequence. The dozen or so candidate main-sequence binaries, Iying quite close to this line, are readily apparent.
HIGH-ENERGY ASTROPHYSICS 279 result from an extensive series of eclipse observations would probably not be informative. Radial velocity variations, on the other hand, are certainly present if the objects are truly binanes, and, as always subject to the un- known inclination factors, will be substantial in the close binaries that we seek Of course, with our dozen or so candidates, it is quite unlikely that all have unfavorable inclinations. In collaboration with M. Bolte of the Dominion Astrophysical Obser- vatory, we have begun a program to obtain repeated observations of radial velocities for the ten or so brightest, best separated objects on this parallel man-sequence in NGC 288. This is a difficult observational problem: we seek a large number of velocities of accuracy ~10 km s-i on objects of ~ > 19.5. Until very recently, one would be hard pressed to find in the lit- erature a spectrum of any cluster main sequence star, much less a program of repeated observations of many objects at reasonably high dispersion. The recent development in instrumentation that has made this program feasible is the introduction of multiple-object fiber-fed spectrographs at several of the world's 4m-class telescopes. This project is of course ideal for such an instrument, as all of the candidate stars are in immediate proximity, lying on the same 3' x 5' CCD chip, and we desire all of the radial velocities simultaneously, not knowing which are true binaries, and which have favor- able inclinations. Further, we can employ the many otherwise unused fibers to obtain spectra of brighter cluster members in the immediate proximity, which are not only of astrophysical interest, but provide a large number of local velocity standards,2 as well as a varieW of control objects so important to this precision work, e.g. anonymous stars of comparable magnitude and crowding to our candidate binaries. The colors of the candidates indicate spectral Apes near solar, and so we may anticipate ample absorption lines for radial velocity determination. We are conducting our observations using the fiber-optic coupled multi- object spectrograph (FOCAP) (Gray 1986; Party and Gray 1986) at the 3.9m Anglo-Australian Telescope. This system uses approximately 60 fibers to feed as many star or sly spectra to the no' offal RGO spectrograph, using the Image Photon Counting System (IPCS) as a two-dimensional detector. We use a grating which yields ~1000A of spectral coverage, with 2~ resolution, centered near A4200 to take advantage of the region of peak efficiency of the IPCS. We have thus far obtained observations at two epochs in 1987 88 which convince us that the project is Just!) feasible in excellent observing conditions; typical integration times to obtain ~100 counts pixels above sly at this dispersion on these faint stars are 12,000 sec. In Figure 2, we show an example of one of the spectra so obtained of one of our main 2 Several dozen giants and horizontal branch stem in NGC 288 already have velocities determined to an accurapyof ~.6 k n s 1 via radialvelocity spectrometer observations (P~yoretaL 19~.
280 AMERICAN AND SOVIET PERSPECTIVES 6000 4000 an o 2000 o 100 50 o T I ~I T I l I AL80 I I ~1 11 4000 4500 Wavelength (I) 1 11 1 1 1 1 ,.l - uB1 4000 4500 Wavelength (I) FIGURE 2 To of the sixty spectra obtained simultaneously in the NGC 288 field at the Anglo-A~lstralian Telescope in 1988 July, using the fiber-optic coupled multi- object spectrograph (FOCAP) and the Image Photon Counting System (IPCS) detector. Lower panel: One of the candidate I-sequence binaries, MB1, an object with V = 19.2, (B-V) = 0.50; despite the modest signal-to-noise ratio of the spectrum, one can clearly see Ht. H7, and Ca H & K absorption lines. Upper panel: one of the bright radial velocity standard stars in the cluster, AL80, a giant with V = 13.89, (B-V) = 1.14 (AIcaino and Liller l980~; a precise (~0.7 km s-1) velocity for this star has been obtained by Pryor et aL (1988~) with a radial velocity spectrometer. All of the features and structure in this spectrum are actual absorption lines in the object. At this dispersion, cross~orrelation of the two stars can yield velocities of accuracy ~6 km s-1 if there is reasonable signal in the program objects
HIGH-ENERGY ASTROPHYSICS 281 sequence binary candidates, together with that of one of the bright giant radial velocity standards in the cluster. Cross-correlation analysis allows us to derive velocities for the binary candidates to an accuracy of ~6 lain sol. As the necessary multiple-epoch observations are still under way, these results must be viewed simply as a progress report at this time. However, for the candidates where we have sufficient signal, we can already say, based on their derived velocities, that these faint stars are indeed cluster members; their sing position on the C-M diagram cannot be explained merely by superposition of a non-member. Thus we are encouraged by the prospect that current instrumentation is indeed adequate to reveal radial velocity variations in main sequence binaries in this and similar clusters in the near future. CONCLUSION In an lateral of only a few years, we have progressed from knowing of literally no non-degenerate binary stars in globular clusters, to a short list of candidates of a variety of luminosities. Whether the very crowded inner cores of the clusters currently hide a population of close binaries that would otherwise be easily distinguished by their odd colors and luminosities will presumably become clear with even the first handful of multicolor images from the Hubble Space Telescope. Thus prospects are bright that in the next few years, observations will finally catch up to, and hopefully overtake, the theory of close binary stars in globular clusters. ACKNOWLEDGEMENTS Most of what little I know about globular clusters and binaries therem has been taught to me by Michael Bolte, and I am indebted to him for his continued collaborative efforts in much of the work described here. I thank Russell Cannon and all of the staff of the Anglo-Australian Observatory for hosting me during the sabbatical year where much of this work was done; Warrick Couch was particularly essential in accomplishing many of these observations. I appreciate useful discussions with P. Hut, D. Popper, C. Pryor, and H. Richer. E. ballots has aided in the data reduction. I was a Visiting Astronomer at Cerro lblolo Interamencan Observatory, National Optical Astronomy Obse~vatones, which is operated by AURA, Inc., under contract to the NSF. This work has been supported in part by NASA Contract NAS5-29293 and NASA Grant NGT-70050. REFERENCES Alcaino, G., and W. tiller. 1980. The main sequence of the globular cluster NGC 288. Astronomical Journal 85: 159~1603.
2~2 AMERICAN AND SOVIET PERSPECTIVES Alcaino, G., and W. Liller. 1987. BVRI CCD photometry of Omega Centaun. Astronomical Journal 94: 158~1599. Auriere, M., Lo Koch-Miramond, and S. Ortolani. l9S9. Ibe X-ray source in the core of 471hcanae. Astronomy and Astrophysics 214: 11~122. Bailey, S. I. 1902. A discussion of the variable stars in the cluster co Centaun. Harvard Annals 38: 1-246. Bell, R A, G. ~ H. Harris, J. E. Hesser, and R. D. Cannon. 1981. Spectroscopic evidence for a wide range in abundances among faint subgiant stars in the globular cluster Omega Centaun. Astrophysical Journal 249: 637 646. Bolte, M. 1987a. Main sequence CCD photometry of the globular duster NGC 362. Astrophysical Journal 315: 469~79. Bolte, M. 1987b. Deep CCD photometry of the globular cluster NGC 7099. Astrophysical Journal 319: 760 771. Bolte, M. 1989. Mass segregation in the globular cluster M30. Astrophysical Journal 341: 16~174. Cannon, R. D., and R. S. Stobie. 1973. Photometry of southern globular clusters. I. Bright stars in w Centauri. Monthly Notices of the Royal Astronomical Society 162: 207-225. Cordova, F. A., and K O. Mason. 1983. Accreting degenerate dwarfs in close binary systems. Pages 147-187 In: Lewin, W. H. G., and E. P. J. van den Heuvel (eds.~. Accretion Dnven Stellar X-ray Sources. Cambridge University Press, Cambridge. Cordova, F. A., and K O. Mason. 1984. X-ray observations of a large sample of cataclysmic variable stars using the Einstein Observatory. Monthly Notices of the Royal Astronomical Society 206: 879~97. DaCosta, G. S., J. Norris, and J. V. V0lumsen. 1986. The blue straggled of Omega Centauri. Astrophysical Journal 308: 74~754. Dickens, R. J., I. R. Brodie, E. ~ gingham, and S. P. Caldwell. 1988. A catalogue of magnitudes and coloum in the globular cluster Omega Centauri. Rutherford Appleton Laboratory Publication RAL 88 004. Elson, R., P. Hut, and S. Inagaki. 1987. Dynamical evolution of globular clustem Annual Review of Astronomy and Astrophysics 25: 565 601. Fernie, J. D., and W. J. Rosenberg. 1961. The ejects of unresolved binaries in three~olor photometry. Publications of the Astronomical Society of the Pacific 73: 259~263. Geyer, E. H. 1967. Eine dreifirbenphotometne van Omega Centauri (NGC 5139~. Zeits~ift fur Astrophysik 66: 16-3Z Geyer, E. H. 1971. A photoelectric investigation of the eclipsing binary V78 in Omega Centaun (NGC 5139~. Pages 235-237 In: Proceedings of the IAU Colloquium 15, flew Directions and New Frontiers in Variable Star Research, Veroffentlichungen der Remeis-Sternwarte Bamberg, IX, No. 100. Geyer, E. H., and N. Vogt. 1978. On the membership of the RR5 star V65 and the EA binary V78 in Omega Centauri (NGC 5139~. Astronomy and Astrophysics 67: 297-299. Gray, P. M. 1986. Anglo-Australian Observatory fibre system. Proceedings of the Society of Photo-Optical Instrumentation Engineem 627: 96 104. Grindlay, J. E., P. Herd, J. E. Steiner, S. S. Murray, and ~ P. Iightman. 1984. v ~ Determination of the mass of globular cluster X-ray sources. Astrophysical Journal (Letters) 282: L13-16. Gunn, J. E., and R. F. Gnffln. 1979. Dynamical studies of globular clustem based on photoelectric radial velocities of indMdua1 stars. I. M3. Astronomical Journal 84: 75~773. Har~ick, F., A. Cowley, and J. Gnndlay. 1982. Evidence for extended X-ray emission from globular clustem Astrophysical Journal (Letters) 254: L11-13. Hell, B. E. 1987. Four color photometry of eclipsing binaries. XXVI A. RY Aqr a low-mass semidetached Stem with intrinsic vanability. Astronomy and Astrophysics 172: 15~166. Hertz, P., and J. E. Gnndlay. 1983a. X-ray evidence for white dwarf binaries in globular clusters Astrophysical Joumal (Letters) 267: L83~7.
-~- # ~ 1 ~ ~ 1~. ~ ~ ~ ~ go dog lump Mention. ~mp~i=1 kumal go: 10~19. He=, it, and ~ S. I. 19~. We nalum of the 1~-l~in~ globular duster <~ _ ~1 ~ ~: ~1~. 1 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Sap ~ ~ B. Stern. 1~. ^ ~ ~lo~magnilude slug of ~ Togae. Pupations of We ~mno~1 Stem ~ the PadOc By: 7~8. Hear, 1. E., R. O. Value, ~ ~ H~eo, R. D Orson, ~ ~n Ludlow B. aged and D. Ego. 1~. ^ Day ~or_~ilude diadem far me Door star dialer ~=~1~7~ PI of He ~1~DO~ Idea of the Parsec ad: ~1& Hills, J. G. 1~. En~unle~ Open away and single ~ and 1beir egg on He Swamis column ~ Cellar plea. ~tmnomi=1 loop ~ I. Hogs, H. 3. 1~3. ^ laid Bulge of adage sag iD globular Cusps ~mpd~Dg ~19 =~ PI ~~ ~ 0~ 0~_ 3, ha. ~1~. flogs, H. S., and ~ Pablo. 1~. ^ pbologmpbic n~ in the glower clever flier 1! frugal of the ~1 ~-nomi=1 S-eV ~ 0~ S~ 16~1~. ape, E 1~. Owes and solar ~ludon. 1 Oslo Bulledn ha: 91. ]en~n, ~ S., and H. E. Lena. 19~. Cog band ~ and ~ ligb~u~= far He echoing gnaw ~ ~ in Omega redbud. Logo and ~l~p~io 3upp~ment ha. ~1~ __ -mood, lo, and at. Adam. 1~7. <~ and ~ Option of w proud E=~ ~l~no~ and ~lmp~o 183: 1~. # ~ 1~ ~ ~ -~ ~ b ~ ~ _ ~ _ ~ Doper ~= _~1 ]-m~ Ha: Id. Pages 31~0 In: Aped D., and B. Shadow Ada. Globular Ousle=. Om~dge ~9 P_ ~ In, ~ H. G., and ~ C Ha. 1~. K~ buses and the X~ Cup of He _~ _~ ~>~R1 ~ Dam Sap ~ 3~. Om~- Uncap Pap Samba. Odes M. H. 1~ ~ flag binary iD He Beld of w Am. I^~ali~ Bulletin on adage Sag 15~. = E. 1~. gong On =~ndedi~en slemeD. amiss Den as: a. In, ~ ~ amp, ad 0. I. 1~. We ~ Hem ~ phi a~ is ligb1~. ~tm~io and Saw Science 47: I. Carol, B., and at. B~1= 1~. We l~lumin~iV X- Cup in Omega Into. ~1 amp ~ Ha: ~~on, B., and R. Onn=. 1~) ~ aiming Nue simpler in ~ Maraud. 0~1 1~ Ha. Bags, B., and ~ ~ Dana. lS3. ^ Hand =~lc gamble in a globular gales ~mp~~1 puma -~- Ha: ~1~. If=, B., R. A. Dane, and 1. a. Ou=. 1981 ~ V101: a dam ~ Cam globular deep ~1~ ]~mal ~1~ ad: ~ Adieu, ~ D., and D. ~ Tam. 1~. We s~1ia1 disldbution of Epic Ins ad blue Plea in Ha. Ages 67~76 In: Odadl~ ~ H. O., and ~ 0. D^ Philip beak. P ~ of IT -~ slum 1~, globular Ouster Xylems flies. ~= D0~L Yucca, ~ H. 1~. Lends ~n-=quen~ of Some stellar tousles Only NoU=s of I__ 1~. ~an, O., ad at. _ 1~. Sages of ~am1~1 pappies of globular dales. II. ~ mud=, =1~9 Aspen and mad ~ ~ redbud ad ~ ~. ^1mno~ ad ~l-~= lag 1~= _T ~ ~ C ~ 1~ ~ _~ ~ ~ _r~ ~1 puma ~& 1~.
284 AMERICAN AND SOVIET PERSPECTIVES Nemec, J. M., and J. G. Cohen. 1989. Blue straggler stars in the globular cluster NGC 5053. Astrophysical Journal 336: 780 797. Niss, B., H. E. Jorgensen, and S. Laustsen. 1978. A search for new variables in the globular cluster Omega Centaun. Astronomy and Astrophysics Supplement 32: 387-393. Norris, J. 1980. The correlation of cyanogen, calcium, and the heavy elements on the giant branch of Omega Centaun. Pages 113-123 In: Hanes, D., and B. Madore (eds.~. Globular austere. Cambridge University Press, Cambridge. Party, I. R., and P. M. Gray. 1986. An automated multiobject fibre optic coupler for the Anglo-Australian Telescope. Proceedings of the Society of Photo-Optical Instrumentation Engineers 627: 118-124. Patterson, J. 1984. The evolution of cataclysmic and low-mass X-ray binanes. Astrophysical Journal Supplement 54: 443~93. Pogson, N. 1860. Remarkable changes observed in the cluster 80 Messier. Monthly Notices of the Royal Astronomical Society 21: 3~33. Pryor, C. P., D. W. Latham, and M. L Hazen. 1988a. A search for spectroscopic binaries in the globular cluster M3. Astronomical Journal 96: 12~138. Pryor, C., R. D. McClure, J. M. Fletcher, and J. E. Hesser. 1988b. A survey of globular cluster velocity dispersions. Pages 661 662 In: Grindlay, J. E., and A. G. Davis-Philip (eds.~. Proceedings of IAU Symposium 126, Globular Cluster Systems in Galaxies. Kluwer, Dordrecht. Pryor, C., R. D. McClure, J. E. Hesser, and J. M. Fletcher. 1989. The frequency of primordial binary stars in globular clusters Pages 175-181 In: Merritt, D. (ed.~. Dyanamics of Dense Stellar Systems. Cambridge University Press, Cambridge. Richer, H. B., and G. G. Fahlman. 1987. Deep CCD photometry in globular clusters V. M5. Astrophysical Journal 316: 18~205. Shara, M. M., J. Kaluzny, M. Potter, and A. F. J. Moffat. 1988. A CCD search for faint variables in the field of an w Centauri low-luminosity X-ray source, and in 47 1hcanae. Astrophysical Journal 328: 594 599. Sham, M. M., ~ F. J. Moffat, and M. Potter. 1987. Outburst and quiescence observations of the dwarf nova V101 in the globular cluster M5. Astronomical Journal 94: 357-359. Shara, M. M., A. F. J. Monet, M. Potter, H. S. Hogg, and ~ Wehlau. 1986. Flmt optical candidate for a recovered classical nova in a globular cluster: Nova 1938 in M14. Astrophysical Journal 311: 796 799. Sistero, R. F. 1968. The eclipsing binary V78 in Omega Centaun. Information Bulletin on Vanable Stars 316. Steinlin, U. 1956. Zur Anwendung der dreifarbenphotometrie in der stellarstatistik. Zeitschnft fur Astrophysik 39: 21~}218. Sttyker, L L, and B. J. Hrivnak. 1984. A search for radial velocity variations in the blue straggler of NGC 7789. Astrophysical Journal 27& 21~219. Verbunt, F., and G. Meylan. 1988. Mass segregation and formation of X-ray soured in globular dusters. Astronomy and Astrophysics 203: 297-305. Verbunt, F., R A. Shafer, F. Jansen, K A. Arnaud, and J. van Paradijs. 1986. A soft X-ray observation of ~ Centaun with EXOSAI: Astronomy and Astrophysics 168: 16~172. Verbunt, F., J. van Paradijs, and R. Elson. 1984. X-ray sources in globular clusters Months Notices of the Royal Astronomical Society 210: 899 914. Walker, A. R 1986. CCD photometry with small telescopes. Pages 33~6 In: Hearnshaw, J. B., and P. L" Cottrell (eds.~. Proceedings of the LOU Symposium 118, Instrumentation and Research Programmes for Small Telescoped Reidel, Dordrecht. Wheeler, J. C 1979. Blue stragglers as long-lived stars. Astrophysical Journal 234: 56~578. Woolly, R v. d. R 1966. Studies of the globular cluster ~ Centauri I. Catalogue of magnitudes and proper motions Royal Obsenratoty Annals 2: 1-128.