1. Later observations did not indicate that there was any need to change these elements. (W.E. Harper, Pub. DAO, 6, 208, 1935).

2. The orbit was assumed to be circular, and the orbital elements are based on measures of two lines of He II seen in absorption. The velocities show a large scatter, and Hutchings offers several alternative solutions based on different emission lines and different numbers of plates. The system resembles some X-ray binaries and may contain a collapsed component. The primary generates a strong stellar wind, and it is unclear how far any of the spectral lines represent the true orbital velocity of the star. P.S. Conti & J.-M. Vreux (ApJ, 228, 220, 1979) do not confirm the period found by Hutchings and question whether the star's velocity is genuinely variable.

3. The new elements, based on photoelectric velocity measurements, should certainly be preferred over those originally determined by W.H. Christie (ApJ, 77, 310, 1933), especially since the new observations lead to an improved value of the period. The agreement between the two values of K is acceptable. Beavers and Salzer applied the test devised by Lucy & Sweeney to the eccentricity and concluded that, although small, it is genuine. The star varies in light by about 0.2 mag in V in a way that suggests it is an eclipsing variable. (R.D. Lines & D.S. Hall, (IBVS, No. 2013, 1981). The shape of the light curve suggests that at least one component is distorted - unusual in a system of such long period. Attempts to detect the secondary spectrum have so far failed.

4. Earlier investigations by R.H. Baker (Pub. Allegheny Obs., 1, 22, 1908), H. Ludendorff (ANac,178, 23, 1908), O. Kohl (ANac,262, 472, 1937), W.J. Luyten, O. Struve and W.W. Morgan (Pub. Yerkes Obs., 7, Pt. IV, 1, 1939) and J.A. Pearce (Pub. AAS, 9, 16, 1936) gave a range in K from 27 km/sec to 34 km/sec. Aikman obtained new, high-dispersion observations of this mercury-manganese star; the orbital elements given are based on his new observations and the older ones. No new elements have been published since, but K.D. Rakos, H. Jenker and J. Wood (A&AS, 43, 209, 1981) find evidence for light variations in the ultraviolet with a period of about 0.96 and velocity variations (measured from three Si II lines near = 1300Å) of up to 10 km/sec in a period of about 0.13. S.J. Adelman and D.M. Pyper (A&A, 118, 313, 1983) also suspect some light variations at visual and near ultraviolet wavelengths. Considering these results, and a possibility that V may be varying, we believe that the assessment given for this orbit in the Seventh Catalogue may have been optimistic. (Note also a misprint in Aikman's paper where 9.6 km/sec is given as the most recent value of V rather than -9.6 km/sec.) Petrie's (II) value of m = 1.35 is probably affected by the 4479 component of the Mg II line and is an underestimate. The star is the brighter component of ADS 94; the companion is 11.4 at 75".

5. The orbital elements are described as `preliminary' by Hube and Gulliver themselves. Curchod and Hauck classify the metallic-lined spectrum as A3 from the K line and F0 from the metal lines. Hube and Gulliver suggest that a search for eclipses might be worthwhile.

6. The new results by Andersen et al. represent a major advance in our understanding of this system. The velocity-curve of the secondary component (the K star) is now well determined by photoelectric measurements - much better so than the d quality suggests. The value of V is derived from measures of the secondary component, which also leaves little doubt that the true orbit is circular (the epoch is the time of primary minimum and the period is variable). Earlier work by O. Struve (ApJ, 99, 89, 1944) and by C.J. van Houten (A&A, 97, 46, 1981) is now superseded. Andersen's own determination of K from spectrograms (PASP, 85, 191, 1973) is surprisingly closely confirmed. There is now general agreement that the A-type spectrum is that of a shell, not that of the hotter star, whose true spectral type was estimated from UV observations to be B7 (M. Plavec, J.L. Weiland and R.H. Koch, ApJ, 256, 206, 1982). Andersen et al. have also analyzed the photometric observations, particularly those of C.-Y. Shao (AJ, 72, 480, 1967). They find an orbital inclination of about 79º and a light ratio in V (K star brighter) of 1.7. The system appears to be semi-detached. E.F. Guinan, S. Tomcyzk and D.J. Turnshek (PASP, 95, 364, 1983) confirm that the H emission comes from a region larger than either star. The faint companion (V = 11.68) about 17" away appears not to be physically related to the binary system.

7. This is a cataclysmic variable of the Z Cam type. Even the period is uncertain - an alternative value of 0.149 is quite possible - and the elements must be considered provisional, although K and V would be approximately the same whichever period is chosen. Co-adding individual spectra gives a somewhat lower value of K (92 km/sec) and an appreciably better fit to the velocity-curve. A circular orbit was assumed.

8. This is the first complete investigation of the system since the original one by J.A. Pearce (MNRAS, 92, 877, 1932) although O. Struve and M. Rudkjobing (ApJ, 108, 537, 1948) and J. Sahade (Pub. Goethe Link Obs., No. 69, 1967) questioned the visibility of the secondary spectrum seen by Pearce. In consequence, the very high masses found for this star by the last-named investigator have long been doubted. Hutchings and Bernard base their analysis on only eleven spectrograms, but since these are well distributed around the orbital period and the results for the primary spectrum agree well with Pearce's, we can probably assume that the orbit of the brighter component is well known. They believe that Pearce measured lines of Fe II in the neighbourhood of the K line and misidentified these as the secondary component of that line. They find possible traces of a weak secondary component in the helium lines which indicate a mass-ratio near unity and a magnitude difference (in the photographic region of the spectrum) of about 1.5. Corresponding minimum masses are around 15 M for each component. The epoch is T. The secondary may be of earlier spectral type and still close to the main sequence. There is evidence for a stellar wind associated with the primary star.

9. The spectral type is A2 from the K line and F2 from the metallic lines.

10. An earlier investigation by E. MacCormack (ApJ, 80, 120, 1934) gives elements in good agreement with Cester's except for a small eccentricity. The light-curve obtained by N.L. Magalashvili & Ya.I. Kumsishvili (Bull. Abastumani Obs., No. 22, 3, 1958) shows a displaced secondary minimum which suggests an appreciable eccentricity of the orbit. K.J. Johnston (ApJ, 176, 455, 1971) has confirmed Cester's suggestion that the light variation is due solely to the ellipticities of the two nearly identical components. He found i = 38 and suggested there might be rapid apsidal rotation. New spectroscopic observations are being obtained by C.T. Bolton to test this possibility. The epoch given in the Catalogue is T. The star is the brighter component of ADS 191; companion 7.8 at 12".

11.

12. Earlier investigations were made by W.S. Adams and G. Stromberg (ApJ, 47, 329, 1918), J.A. Pearce (Pub. DAO, 3, 275, 1926), and O. Struve and H.G. Horak (ApJ, 110, 447, 1949). All available observations have also been discussed by G. Mannino (Asiago Contr., No. 103, 1959) and A. Krancj (Pub. Bologna Univ. Obs., 7, No. 14, 1960). An excellent three-colour UBV light-curve was obtained by R.H. Koch (AJ, 65, 127, 1960). These and an earlier light-curve obtained by F.B. Wood (ApJ, 108, 28, 1948) were analyzed by J.B. Hutchings and G. Hill (ApJ, 167, 137, 1971) by their method of light-curve synthesis. They found the orbital inclination to be 56, the eclipses being grazing and most of the light variation being the result of distortion of the stars. Abhyankar finds evidence for different mean velocities for each component, raising the question how far the velocities derived for the secondary star can be assumed to result from its orbital motion. If the secondary spectrum does arise from the secondary star then the system does not conform to the mass-luminosity relation. Petrie(II) found m = 0.78, a greater difference than would be expected if the mass-ratio is 0.85, as found by Abhyankar. Hutchings and Hill confirm this value of the mass-ratio (they find 0.81) and speculate that the system is close to a rapid phase of mass transfer. The orbit of the primary is well determined in Abhyankar's work, although his value of K is different from those found earlier. He believes there is some evidence for apsidal rotation with a period of about 70 years.

13. Light-curves in B and V have been published by R.M. Williamon, T.F. Collins and K.-Y. Chen (A&AS, 34, 207, 1978). The epoch is the time of primary minimum. Hilditch and King were unable to satisfy the light-curves with the same value of the mass-ratio as obtained from the velocity-curves; the light-curves give q = 0.27. Subject to uncertainties arising from this, Hilditch and King find the masses, radii and luminosities of the two stars to be, respectively: 1.9 M and 0.7 M; 2.1 R and 1.3 R; and 8.95 L and 3.47 L. The orbital inclination is close to 80. Different luminosity ratios would be found at the two quadratures from the relative intensity of the spectra.

14. Although the observations are spread over an interval of nearly sixty years, no evidence for apsidal motion has been detected. The original paper gives spectrophotometric information and an estimate of rotational velocities (75 km/sec). The study serves as an example of what can be learned when old plates are measured by modern cross-correlation methods.

15. It is, perhaps, surprising that this relatively bright, short-period system has not attracted more attention from spectroscopists. Plaskett's discussion remains the only complete one. Luyten rediscussed the observations and derived a small eccentricity: photometric observations, however, reveal that any displacement of the secondary minimum is very small. Observations by J. Sahade and O. Struve (ApJ, 102, 480, 1945) agree with Plaskett's curve but show a distinct rotational disturbance. Those investigators could not confirm Plaskett's detection of the secondary spectrum (K = 150 km/sec) - hardly surprising since modern photometry indicates a light-ratio in V of about 0.05 (K. Wlodarczyk, Acta Astr., 34, 47, 1984). Plaskett's zero of phase is based on a time of primary minimum given by R.J. McDiarmid (Princeton Obs. Contr., No. 7, 1924). The value given in the Catalogue is one-quarter of a period later and corresponds to the time of maximum positive velocity. There is evidence, however, that the period changes (A.C. de Landtsheer, A&AS, 52, 213, 1983). De Landtsheer also presents an infrared light-curve in that paper. In addition, with P.S. Mulder (A&A, 127, 297, 1983) he reports on IUE observations of the system, which failed to show any trace of circumstellar matter. A discussion of the system's possible evolution has been published by J.P. de Greve, A.C. de Landtsheer & W. Packet (A&A, 142, 367, 1985). Wlodarczyk (op. cit.) finds i = 80, in good agreement with earlier discussions of the light-curve (C.M. Huffer & Z. Kopal, ApJ, 114, 297, 1951) and J. Papousek, Bull. Astr. Inst. Csl, 25, 152, 1974). The spectral type given for the secondary component is derived from Wlodarczyk's solution.

16. Original observations and elements by G.H. Tidy (Pub. DDO, 1, 191, 1940). Tanner showed that the period adopted by Tidy was incorrect. The epoch is T as defined by Sterne. Lucy & Sweeney regard the orbit as circular.

17. Period uncertain by up to one day. Epoch is arbitrary zero of phase. Star is brighter component of ADS 328: companion 8.0 at 0".2, but star appeared single in 1953 (IDS).

18. Secondary spectrum seen and measured with difficulty.

19. Hube calls attention to the large mass-function and suggests that the system may prove to be eclipsing. Earlier reports of variability, however, have not been confirmed. Some authors have classified the star as an Si Ap star, but Hube finds no evidence of peculiarity. J. Zverko (Bull. Astron. Inst. Csl., 30, 372, 1979) published an abundance analysis and concluded that silicon is slightly overabundant.

20.

21. Residuals are small, but maximum of velocity-curve is not well defined. H.L. Alden (Pub. AAS, 9, 31, 1937) gives an astrometric orbit, i = 110.4, a = 0".068, = 3.

22. The orbit is based exclusively on photoelectric measurements of radial velocity, although earlier photographic measurements revealed the variability (R.E. Wilson & A.H. Joy, ApJ, 111, 221, 1950); J.F. Heard, Pub. DDO, 2, 105, 1956)). A circular orbit was adopted after an application of Bassett's test (Observatory, 98, 122, 1978) showed the small eccentricity to be statistically insignificant. The epoch is T. Griffin suggests that the star may be fainter than the H.D. magnitude given in the Catalogue.

23.

24. This Cepheid variable is also a spectroscopic binary and a member of the cluster NGC 129.

25. This is another supergiant member of the cluster NGC 129.

26. Star is 1' S.E. of B.D. +61 113. Period is accurately known from light-curve. I.M. Levitt (Flower Reprint, No. 76, 1949) found from visual observations i = 67 and the light-ratio to be about 0.75.

27. 13 Cet (ADS 490) is a well-known triple system. Luyten estimates V for the whole system is 11.0 km/sec. Earlier investigations are by J.S. Paraskevopoulos (ApJ, 52, 110, 1920) and A. Pogo (ApJ, 68, 116, 1938). A visual orbit for the long-period system (P = 6.91) was derived by Aitken (Pub. Lick Obs., 12, 5, 1914) and rediscussed by Luyten in the paper cited in the Catalogue. The long-period system has not been included in the Catalogue since there seems to have been no direct spectroscopic determination of K, the semiamplitude of the spectroscopic binary about the centre of mass of the whole system. There are discrepancies in the orbital elements from epoch to epoch, probably indicating that the two orbital motions have not yet been completely disentangled, although there may also be physical perturbations. The set of elements given here refers to the interval 1923--28. G. Gatewood and S. Sofia (PASP, 88, 50, 1976) find that the astrometric data suggest that the principal components of the system are overluminous for their masses.

28.

29. This is a mercury-manganese star for which Stickland and Weatherby give a `possible orbital solution'. A periodicity of about 400 in the velocity had previously been suggested by G.C.L. Aikman (Pub. DAO, 14, 379, 1976).

30. This star was used as a reference star in the Cambridge photoelectric velocity measurements and therefore, as explained by McClure et al., some orbits derived from such observations may be affected. It is difficult to classify the orbit. The scatter of observations at a given phase is often greater than the total range of variation, but the large number of observations permits a fairly precise determination of K. In view of the long (and correspondingly uncertain) period, a conservative assessment seems appropriate.

31. Brighter component of ADS 513, companion 9.0 at 36". Other investigations by F.C. Jordan (Pub. Allegheny Obs., 2, 45, 1910) and W.J. Luyten, O. Struve and W.W. Morgan (Pub. Yerkes Obs., 7, Pt. IV, 254, 1939). The elements of the primary orbit found by these authors agree well, and they must be considered well determined. Only Pearce observed the secondary spectrum. Petrie(II) found m = 3.17. Luyten, Struve and Morgan revised the period to 143.621.

32. The Durchmusterung number is from the C.P.D. Radial velocities for this early-type near-contact system have been determined by cross-correlation and the primary curve is well covered although the r.m.s. error is 8.5 km/sec. The epoch is a time of primary minimum, but the period is decreasing and a quadratic term (9.04 × 10) must be included in the ephemeris. Hilditch and King have combined their spectroscopic observations with photoelectric light-curves observed by J.V. Clausen and B. Grønbech (A&AS, 28, 389, 1977) to obtain masses of 1.6 M and 0.7 M, radii of 1.6 R and 1.0 R and luminosities of 5.6 L and 0.36 L. The orbital inclination is close to 81. The secondary spectral type is estimated from the effective temperature quoted by Hilditch and King.

33. Residuals from velocity-curve are small, but curve is based on only eleven observations. On basis of assigned spectral types, m 0.6. Epoch is T. Star is listed in IDS: companion 8.6 at 330".

34. Bakos gives a period of 15,000 days, which can be only an approximate value. The observational data for this low-amplitude binary are taken from several different observatories, and a homogeneous series of observations would be very helpful in confirming the orbital elements. The star is the brighter component of ADS 548: companion is 12 at 28" separation.

35. The velocities are determined by cross-correlation. Emission is seen at H and K, varying approximately in phase with the more massive component. The epoch is the time of primary minimum, when the less massive star is eclipsed. An unfiltered photoelectric light-curve has been observed by D.H. Bradstreet (AJ, 86, 98, 1981) who finds i = 81 and that the brighter (but marginally cooler) component gives 60 percent of the light.

36. The variation of the velocity of this star seems to be established, but with only one observation on the descending branch of the velocity-curve, the period must be uncertain.

37. Orbital elements have previously been published by E.A. Vitrichenko (Bull. SvA, 11, 898, 1968) and Izv. Krym. Astrofiz. Obs., 39, 63, 1969)). The new elements are preferred because they are based on more, better distributed observations. The main difference is the somewhat larger value of K found by Vitrichenko. Gies and Bolton suggest that the secondary is cooler than the primary and overluminous for its mass - they estimate the visual magnitude difference between the two components to be close to zero. Vitrichenko observed a light variation of about 0.2 in V, consistent with the star being an ellipsoidal variable. The epoch is the time of inferior conjunction of the bright star.

38. This system resembles YY Gem, although no eclipses have been detected. Both components have emission of variable intensity in the H and K lines of their spectra. At least one of the stars is a flare star and a BY Dra variable. A model for the system containing a spotty star is proposed by Bopp and Fekel. The epoch is the time of conjunction with the velocity of the primary star increasing: the orbit was assumed circular. The star varies by about 0.06 in V.

39. Earlier investigations by W.E. Harper (Pub. DAO, 4, 135, 1917) revised in (Pub. DAO, 6, 107, 1935), and by Luyten, based on Harper's material. Agreement with new elements is fairly good. Harper found e = 0.01; Mannino and Grubissich find e 0.005. Values of K found by Harper are each about 3 km/sec less than values in the Catalogue. Epoch is T. A further discussion of the system was made by A. Krancj (Pub. Bologna Univ. Obs., 7, No. 11, 1959). Petrie(I) found m = 0.25.

40. Earlier investigations by J. Lunt (Cape Ann., 10, Pt. 7, 38G, 1924) and by Neubauer himself (PASP, 41, 371, 1929). Neubauer's later results agree much better with Lunt's than do his 1929 results. A small systematic difference exists between the Lick and Cape measures. From the Cape measures alone, V = +13.6 km/sec.

41. Detection of the secondary component with a Reticon, combined with high-precision measures of the primary, bring the spectroscopic observations of this system to the high standard that photometric observations set long ago. The new elements for the primary star agree well with those found by C.L. Perry and S.N. Stone (PASP, 78, 5, 1966) and the results from the original study by J.S. Plaskett (Pub. DAO, 3, 248, 1926) agree within their uncertainties. The new observations lead to a circular orbit (e = 0.000±0.003); the epoch is the time of primary minimum. Lacy has rediscussed the excellent light-curve by G.E. Kron (Lick Obs. Bull., 19, 59, 1939) and finds i = 88.3, masses of 2.31 M and 1.35 M, radii of 2.53 R and 1.35 R and m = 2.9. A.C. de Landtsheer and P.S. Mulder (A&A, 127, 297, 1983) report from IUE observations that iron is overabundant by a factor of six in this system. De Landtsheer and J.P. de Grèeve (A&A, 135, 397, 1984) discuss the possible evolution of the system. Star is brighter component of ADS 624: companion is 9.7 at 36".

42. The first suggestion that this symbiotic star might be a spectroscopic binary with P 470 was made by S.E. Smith (ApJ, 237, 831, 1980). The value of the period was assumed by Oliversen et al. A circular orbit was assumed and phases were computed from the time of maximum equivalent width of the H emission, J.D. 2,443,200.5. Different methods of reduction lead to rather different values of K, which quantity is, therefore, uncertain. R.E. Stencel (ApJ, 281, L75, 1984) finds evidence from IUE observations for an eclipse of the hotter star by the cooler. Early reports of large variable magnetic fields in this system (H.W. Babcock, PASP, 62, 277, 1950) are not confirmed by more recent observations (M.H. Slovak, ApJ, 262, 282, 1982). The system is also discussed by M.R. Garcia (AJ, 91, 1400, 1986).

43. The scatter of observations about the mean curve is large, and the possibility of other orbital periods should be investigated. Star is brighter component of ADS 622: companion 11.2 at 33" appears to share the proper motion of the primary.

44.

45. Earlier investigations by J.B. Cannon (Pub. DAO, 2, 143, 1915), H.S. Jones (Cape Ann., 10, Pt. 8, 35, 1928) and Luyten (based on both these sets). Except for systematic differences in the Ottawa observations all investigators agree well. E.M. Hendry (PASP, 92, 825, 1980) found that new observations required no modifications to Gratton's elements. The spectrum displays (possibly variable, see Hendry, loc. cit.) Ca II emission. A small light variation was interpreted by C.M. Huffer (Pub. Washburn Obs., 15, 29, 1928) as a combination of ellipticity effects and shallow eclipses. A UBV light-curve has been published by T.S. Belyakina, V.I. Burnashev and V.M. Zhilin (Izv. Krym. Astrofiz. Obs., 56, 16, 1977) who find a range of 0.2 in V and a period of 17.7673. The epoch is the time of inferior conjunction of the visible star. Several faint companions are listed in IDS, but their physical connection is doubtful.

46. The elements given in the Catalogue were published only shortly after those by H.A. Abt and S.G. Levy (ApJS, 30, 273, 1976). Since Nadal et al. use more observations and discuss the system more thoroughly, their results are preferred. The agreement is quite good, but there are some systematic departures of Abt's and Levy's observations from those obtained at Haute Provence. Nadal et al. argue for the existence of a third body causing supposed changes in K, and V. Only the last-named are significant, however, and - despite attempts to connect all observations to the `Lick System' - may be the result of systematic errors between observatories. By Petrie's method, Nadal et al. find m = 0.18 in the photographic region of the spectrum. Two faint companions listed in IDS are regarded by both sets of investigators as probably optical.

47. Abt & Levy combined their new observations with those obtained by F.C. Jordan (Pub. Allegheny Obs., 1, 191, 1910) and offer these new improved elements. Petrie(I) found m = 0.76.

48. Popper notes that the hydrogen lines strengthen with respect to the metallic lines during primary eclipse, but this effect cannot be caused by the secondary spectrum which should be that of a cooler star. There is also an apparent rotational disturbance in the velocities although the lines of the spectrum are quite sharp. All lines except 4481 Mg II, which gives a higher value for K, appear double near the time of primary minimum. The magnitude given in the Catalogue is derived from a single observation. The epoch is the time of primary minimum, and the orbit was assumed circular.

49. The primary component of this two-spectra binary is a newly discovered mercury-manganese star. The velocities of the secondary component were used only to determine K. More recent observations show some systematic departures from the orbital elements given here.

50. Petrie(II) found m = 0.29. The orbit should probably be assumed circular. According to IDS, there is an 11.5 companion at 133".4.

51.

52. Earlier spectroscopic observations are discussed by E.F. Carpenter (ApJ, 72, 205, 1930), Z. Kopal (Harvard Obs. Bull., No. 914, 1950) and R.H. Hardie (ApJ, 112, 542, 1950). The complexity of the spectrum, arising from the presence of circumstellar matter in the system, probably ensures that the orbital elements will never be known with high accuracy. The epoch is the time of primary minimum; the period is variable and increasing. The value of K comes from the discussion of Reticon observations by J. Tomkin (ApJ, 244, 546, 1981). It agrees closely with the value obtained by Batten and is certainly more reliable. Eruptive events observed in 1974 and since (A.H. Batten et al., Nature, 253, 174, 1975) and M. Plavec & R.S. Polidan, Nature, 253, 173, 1975) have stimulated much work on this system. Photometric studies of the variable circumstellar disk have been published by E.C. Olson (ApJ, 237, 496, and 241, 257, 1980) and many of his conclusions have been confirmed and complemented by study of the UV spectrum (M.J. Plavec, ApJ, 275, 251, 1983), whose spectral classifications (similar to Batten's) are given in the Catalogue. Plavec also finds m 2.5. A modern discussion of the undisturbed light-curve was published by E.C. Olson (PASP, 96, 162, 1984) who finds i = 85.8. Polarization in the circumstellar matter has been studied by V. Piirola (A&A, 90, 48, 1980) and A&AS, 44, 461, 1981). Weak X-rays have been detected from this system (N.E. White & F.E. Marshall, SAO Special Report, Vol. 2, No. 392, 93, 1982). These papers form only a part of the recent literature on this active system. Star is brighter component of ADS 830: companion 11.2 at 13".8.

53. This is a Wolf-Rayet binary in the Small Magellanic Cloud, possibly associated with the cluster and H II region NGC 346. The system displays eclipses and the period was determined photometrically (J. Breysacher & C. Perrier, A&A, 90, 207, 1980) and the orbital inclination is estimated to be about 80. Photometric and spectroscopic values of e and are in agreement, but Breysacher et al. point out that the large eccentricity is unusual in a Wolf-Rayet system with only a moderately long period. The epoch is the time of periastron passage as derived from the O-type spectrum. They also find the derived masses to be low, and express doubts whether or not the He II 4686 emission truly represents the orbital motion of the Wolf-Rayet component. That line may also be in emission in the O-type spectrum.

54. One node of this highly eccentric orbit is very well observed. The two components are blended, however, over most of the orbit. Velocities are obtained from 4481 Mg II only. Petrie(II) found m = 0.16.

55. The new data obtained by Hutchings and Thomas supersede those on which Kraft's (ApJ, 135, 408, 1962) original orbit was based, clarifying some issues and confusing others. There appears to have been an error in the tabulation in Kraft's paper, where a value was given for K incompatible with the velocity-curve as drawn. Kraft's data lead to a value of K similar to that found by Hutchings and Thomas. The high orbital eccentricity found by Kraft is probably a result of distortion of the spectral line profiles by the hot-spot. Hutchings and Thomas found a smaller eccentricity than Kraft did and preferred to adopt a circular solution. The epoch is T. There is evidence for a period change between the two epochs of observation. Hutchings and Thomas also believe they have detected the secondary spectrum, which should be measurable in the red region of the spectrum. The system is of the Z Cam type; light variations are not caused by eclipses (P. Szkody, PASP, 86, 38, 1974). Attempts to detect soft X-rays from the system during a flare led only to the setting of an upper limit (P. Henry et al., ApJ, 197, L117, 1975).

56. This is another Wolf-Rayet binary in the Small Magellanic Cloud, first observed by A.F.J. Moffat (ApJ, 257, 110, 1982). The system is not yet known to display eclipses, but Moffat believes that it may be found to do so; he had found a period of 6.861, ruled out by the new observations. The designation signifies the number of the star in Sanduleak's list (AJ, 73, 246, 1968). The epoch is T for the absorption component.

57. The period given in the Catalogue is derived from the radial velocities alone - a somewhat heterogeneous set. The spectroscopic elements are all still highly uncertain, but S.L. Lippincott (ApJ, 248, 1053, 1981) has published a thorough discussion of the astrometric orbit of this Population II binary. Her value for the period of 21.43 is certainly closer to the truth than that derived spectroscopically. She finds i = 109.5, = 335.9 and e = 0.61. She estimates m 4.5. Several faint and distant companions are listed in IDS

58. Bertaud and Floquet list the spectral type as Am, without further qualification. Bennett et al. comment on the prominence of lines of strontium in both spectra. There is no distinguishable difference between the spectra of the two stars.

59. This recently discovered cataclysmic variable is remarkable for its short period and eclipses of a few minutes duration. Orbital solutions are necessarily uncertain. The orbital inclination is probably around 76, the mass of the white dwarf probably between 0.5 M and 1 M and that of the other component probably around 0.25 M. The epoch is mid-eclipse of the `disk'.

60. The new observations by Andersen are of the highest quality and supersede even those of D.M. Popper (ApJ, 162, 928, 1970), with which they agree well except perhaps for V, as well as those of C. Hagemann (MNRAS, 119, 141, 1959) and A. Colacevich (PASP, 47, 84, 1935). Photometric observations (J.V. Clausen et al., A&A, 46, 205, 1976) show e 0.01 and consequently a circular orbit was assumed in a spectroscopic solution. The epoch is the time of primary minimum. The small eccentricity appears to be real, however, and changes in the period suggest apsidal motion. The orbit would be of a quality if these were verified. The light-curve gives i = 88 and L/L = 0.28 (at 5500). There are companion stars each of 7.0 at 0".7 and 6".4 (IDS). The question of their physical relationship to Phe itself is still open, but the spectrum of the closer companion shows up in the combined light of the system.

61. The new study by Andersen et al. supersedes the already good results obtained by M. Imbert (A&AS, 36, 453, 1979) and by B.J. Hrivnak and E.F. Milone (ApJ, 282, 748, 1984). This system is now amongst those with the best determined absolute dimensions, certainly of systems containing an evolved component. The discussion by Andersen et al. is very complete, including atmospheric abundances and evolutionary status. The epoch is the time of primary minimum and the eccentricity (actually 0.188) and longitude of periastron have been constrained to agree with the photometric value of ecos . The orbital inclination is close to 88.5 and the two stars differ by 0.17 in V. Masses are known to better than one percent.

62.

63.

64. This star is a triple system since the spectroscopic pair is one component of the visual binary ADS 999 with a possible orbital period of 75 years. The spectroscopic pair may prove to show eclipses. The spectral type is the mean for all three components (maximum separation in the visual orbit is less than 1"). Duquennoy computes F8 V and F9 V for the two components of the spectroscopic binary and believes that all three stars lie on the main sequence, with the visual companion about F4. O.J. Eggen (AJ, 70, 19, 1965) who did not know of the duplicity of one component, supposed the system to contain evolved stars. The two components of the visual binary are almost equal in luminosity while V for the components of the spectroscopic pair probably lies between 0.2 and 0.4.

65. The new orbit supersedes that by W.H. Christie (ApJ, 77, 310, 1933) both because of the greater precision of the new observations and because they reveal the secondary spectrum. Beavers et al. estimate m = 1.6. They find that the orbital eccentricity, although small, is very probably real. This spectroscopic binary is the fainter member of the common-proper-motion pair that makes up Psc. The brighter, A-type star, Psc A}, is 23" distant and has V = 5.24. A 12.2 component C, about 1" from B, is probably also related. Each of A and B have shown evidence of being double during occultations.

66. A mercury-manganese star for which `possible' orbital elements are given by Stickland and Weatherby. G.C.L. Aikman (Pub. DAO, 14, 379, 1976) also considered the velocity of this star to be variable.

67. The authors offer two solutions, the circular orbit given in the Catalogue and an elliptical one which they prefer on the grounds that most similar systems have elliptical orbits. The eccentricity proposed is 0.16 ± 0.07, and V, K and the standard deviation of an observation of unit weight are almost the same in the two solutions. There seems, therefore, no reason for adopting the elliptical orbit. If the orbit is elliptical, the period would be slightly shorter (11.588); the epoch is T.

68. Unpublished elements provided by Fekel supersede the preliminary ones published by T. Simon, F.C. Fekel and D.M. Gibson (ApJ, 295, 153, 1985). The system appears to be of the RS CVn type. IUE observations reveal the presence of a white-dwarf companion. Radio flares have been observed and the system is a source of soft X-rays (F.M. Walter & S. Bowyer, ApJ, 245, 671, 1981). An 11.2 companion at 177".6 is listed in IDS - only one measurement of its position appears to have been made (by Burnham).

69. This is an X-ray pulsar and the measured quantity is, of course, the delay in the arrival time of the X-ray pulses rather than a radial velocity - which is inferred from the deduced size of the orbit. The systemic velocity is unknown. The elements given in the Catalogue were the first determined. Other investigations by M.J. Ricketts et al. (Space Sci. Rev., 30, 399, 1981) and by R.L. Kelly et al. (ApJ, 251, 630, 1981) lead to very similar elements. The orbit is undoubtedly exceedingly well known, but the existence of apsidal motion is still unsettled. There is some evidence for a slow rate (0.1 yr) of apsidal regression. The system is also known as a -ray source (P.M. Chadwick et al., A&A, 151, L1, 1985). J.B. Hutchings & D. Crampton (ApJ, 247, 222, 1981) have identified an optical counterpart, but were unable to determine any orbital elements for it.

70. Epoch is T. Spectrum very badly distorted by gaseous streams. Modern photoelectric (ubvy I) light-curves have been published by E.C. Olson (PASP, 97, 731, 1985) but no analysis has been attempted except to determine the colours of the components. After allowing for (heavy) reddening, the colours suggest somewhat earlier spectral types (B0.5 and B3) than Struve gave.

71. The elements given here supersede earlier ones by the same authors and some others (F. Primini et al., ApJ, 210, L71, 1976). This is another X-ray binary pulsar, and the orbital elements of the X-ray component are known with very high accuracy. The epoch is the time of minimum delay (or maximum advance) in arrival time of pulses. The eccentricity is known to be <0.0007; the orbital period was fixed in the solution. The values of K and V are derived from optical spectroscopic measurements by J.B. Hutchings et al. (ApJ, 217, 190, 1977) and are much less precisely determined. Hutchings et al. estimate the orbital inclination to be about 70 : Primini et al. confine it between 53 and 73. Hutchings et al. show that He II emission (4686) varies in phase with the X-ray component but shows a somewhat smaller amplitude. D.E. Gruber & R.E. Rothschild (ApJ, 283, 546, 1984) report variability of the X-ray emission and G. Hammerschlag-Hensberge et al. (ApJ, 283, 249, 1984) give the results of ultraviolet spectroscopy. Photometry by J. van Paradijs and L. Kuiper (A&A, 138, 71, 1986) shows the system to be an ellipsoidal variable at optical wavelengths, and the magnitude range given in the Catalogue corresponds to the approximate limits that they found.

72. This is a short-period binary with an active chromosphere. Emission is seen at H and in the ultraviolet. The small light variability is not the result of eclipses but is ascribed by Bopp et al. to starspots. Spectral lines of both components are rotationally broadened and hard to measure. If the components have normal masses for G5 V stars, the orbital inclination is about 30. The absorption lines in the two spectra are described as `of nearly equal intensity'. The epoch is superior conjunction of the more massive star.

73. Earlier investigation by P.D. Jose (ApJ, 114, 370, 1951) was based on an incorrect value for the period. H.A. Abt (ApJS, 6, 37, 1960) classified the spectrum as A3, F2, F5 IV from the K line, hydrogen lines and metallic lines respectively. Fletcher found m = 0.09, by Petrie's method. New observations obtained by Abt & Levy (ApJS, 59, 229, 1985) did not lead them to make any changes to these orbital elements.

74. This star is a visual binary consisting of two nearly equal stars. Fletcher's spectroscopic observations demonstrated that the period was near 16 years rather than 32 years. He gives P = 16.14, T = 1972.742. The elements given are all determined from the spectroscopic observations, although they are close to the latest set of elements derived from the visual observations by W.H. van den Bos (PASP, 74, 291, 1962). The value given for K is K+K and the values of the mass function and asini are correspondingly modified in significance. Slightly different elements were published by C.L. Morbey (PASP, 87, 689, 1975) who has devised a method of combining visual and spectroscopic observations in a single solution for the orbital elements.

75.

76. Earlier investigations by J.H. Moore (PASP, 41, 254, 1929) and B.P. Gerasimovic (ApJ, 84, 232, 1936) are confirmed by Roemer's results. The F8 Ib star is a Cepheid variable with a period of about four days. The magnitude of the star is thus variable through a small range. The secondary spectrum is estimated. There is some slight evidence of a small variation in systemic velocity of amplitude 1 km/sec and period 6 to 8 years, but the available material is insufficient to decide its reality. The velocity variation due to the Cepheid pulsation has a range of 5 to 6 km/sec and is irregular. An astrometric orbit has been published by A.A. Wyller (ApJ, 62, 389, 1957). His value of agrees well with Roemer's and he finds i = 58. The uncertainties, however, are very large. Period is 30.46 and periastron passage is 1928.48. Star is brighter component of ADS 1477, companion 9.0 at 18".

77.

78. Luyten's orbit is based on observations by R.E. Wilson (Lick Obs. Bull., 9, 116, 1917). Wilson assumed a fixed value of T to obtain his elements. Luyten assumed a circular orbit and his epoch is T. D.S. Evans (MNASSA, 16, 4, 1957) on the basis of nine new Cape observations finds the period should be revised to 193.85, but that the other elements need no change. Spectral type is that given by Evans. Star is variable.

79. Elements are based on 22 low-dispersion spectrograms. Light-curves in UBV have been obtained by R.K. Srivastava and T.D. Padalia (Bull. Astron. Inst. Csl., 21, 359, 1970). Eclipses are partial.

80. Star is brighter component of multiple star ADS 1202, which does not seem to be a physical system.

81. The spectral type given is from the HD Catalogue: Griffin and Emerson believe the true spectral type to be closer to K0. They gave T in the modified Julian date system in their paper: 2,400,000.5 has been added to their value to put the time of periastron on the same system as that for other systems.

82. Earlier observations were published by the same author (Acta Astr., 27, 51, 1977). Elements derived from the two sets of observations agree well except for a large change (22.5 km/sec) in the systemic velocity derived from the secondary component. Because of this and the relatively large scatter of the observations, the orbital elements can only be regarded as preliminary. The system is a W UMa system, and the epoch is the time of primary minimum. P.G. Niarchos (Ap&SS, 58, 301, 1978) has derived photometric elements from observations published by R.M. Williamon (AJ, 80, 140, 1975) and finds that i = 82 and the fainter component gives about 0.16 of the light in the yellow region. Duerbeck, however, indicates that the form of the light-curve is variable.

83. This is a Wolf-Rayet binary in the Small Magellanic Cloud. Orbital elements (K and V) are derived from the absorption lines of the O-type spectrum and the O IV emission line (3834) in the W-R spectrum - which gives the most reasonable values for the minimum masses. The epoch is the time of conjunction (O-type star behind) as derived from measures of the absorption lines. The magnitude difference between the components is estimated at 2.6 (O-type star the brighter). Although coverage of the velocity-curve is good, the scatter of individual velocity measurements is very large. Note also the large difference in the systemic velocities derived from the two components.

84. Although no luminosity classification of the spectrum is available, Griffin believes that the star is likely to be a giant.

85. Moffat et al. are very cautious in putting forward this orbit. Velocity variations of the same period are found from the emission line He II 4686 and the absorption line H - but they are displaced in phase by 0.15 from each other. The amplitude is the mean from both lines and the systemic velocity is an approximate value derived from the measures of the emission line (H gives close to -200 km/sec). The magnitude is a photoelectrically determined Strömgren v magnitude. The zero of phase is the epoch at which the He II velocities equal the systemic velocity and are increasing.

86. This star is a long-period eclipsing variable classified by A.P. Cowley (PASP, 81, 297, 1969) as one of the group of systems resembling VV Cep. The orbital elements given here are based on observations by Cowley, J.B. Hutchings and D.M. Popper, and are still only provisional. Nevertheless, the observation of eclipses (J.T. Bonnell & T. Herczeg, IBVS, No. 1146, 1976) leaves no doubt as to the binary nature of the star.

87. Brighter component of ADS 1326, companion 10.6 at 11".

88. Earlier investigations have been published by J.B. Cannon (JRASC, 4, 195, 1910), H. Ludendorff (ANac,186, 17, 1910), J.A. Hynek (Contr. Perkins Obs., No. 14, 1940) and G.R. Miczaika (ZAp, 28, 43, 1950). There is also a thesis by F.R. Hickok from which Poeckert has taken the determinations of the period and zero phase (inferior conjunction of the primary star). The system remains very difficult to interpret, since the presence of shells (probably around both components) creates complex line profiles showing both emission and absorption. Poeckert's study is very thorough and based on excellent material. He has probably determined the orbital elements of the primary star as accurately as is, at present, possible. He himself cautions against assuming too easily that the He II emission line (4686), from which he has determined K, accurately reflects the motion of the secondary star. Masses and dimensions determined for the system do depend on this assumption. Different lines give different values for V.

89. Sterne's method for small eccentricities was used, but epoch is time of periastron passage. Lucy & Sweeney adopt a circular orbit for this system.

90. The optical counterpart of this X-ray source is a star of the AM Her type. The values of K and V given are derived from measurements of the emission lines of hydrogen, ionized helium and ionized calcium (K). The epoch given is the time of inferior conjunction of the source of emission lines, computed from the information given in the paper. The velocity maximum (mean for all lines) occurs at a phase of 0.23.

91. Epoch is T deduced from the photometric data quoted by Struve et al. Photoelectric observations by G.G. Cillié and B.J. Bok (Harvard Obs. Bull., No. 920, 29, 1951) have been analyzed by G. Russo et al. (A&AS, 47, 211, 1982) who find an orbital inclination close to 83, and a fractional luminosity (in V) for the brighter component of 0.6. The two spectra are almost equal in intensity. The system is an X-ray source (R.G. Cruddace & A.K. Dupree, ApJ, 277, 263, 1984).

92. According to Heard and Krautter, D.P. Hube classified the spectrum as B9 III. The star is listed as a mercury-manganese star by Bertaud and Floquet. It has also been studied by G.C.L. Aikman (Pub. DAO, 14, 379, 1976) who obtained very similar orbital elements from all available observations.

93. Because the period is so long, it is not well determined (the Cape observations do not cover one complete cycle, although some Lick observations are available). The orbit has been deemed to be of low quality, although the precision of the individual observations is high. The star is listed in IDS but is an optical pair.

94. For a bright star, this has proved a surprisingly difficult object, mainly - as Pike, Lloyd and Stickland point out - because it is rotating unusually rapidly for a late F-type star. The period has been uncertain. W.E. Harper (Pub. DAO, 3, 113, 1915) first adopted P = 1.73652 which he later revised to 1.73631 (Pub. DAO, 6, 211, 1935). H.A. Abt and S.G. Levy (ApJS, 30, 273, 1976) derived a period of 1.73645. Earlier, Luyten had proposed 2.3413 but R.W. Tanner (Pub. DDO, 1, 473, 1949) showed this to be spurious and even questioned the binary nature of the star. Although the new observations are few in number, they are of good quality and are well represented by the somewhat longer period of 1.767. The r.m.s. scatter of the residuals from the velocity-curve obtained by Pike et al. is several times less than for that published by Abt & Levy. At last, we apparently have reasonably reliable elements for this star. The velocities were determined by cross-correlation and the systemic velocity depends on a conventional measurement of one plate - it is, therefore, much less certain than are the other elements. The companions listed in IDS are probably optical.

95.

96. Magnitude and spectral type are taken from the HD Catalogue since nothing more recent appears to be available. Griffin suggests that the star is probably a giant.

97. A circular orbit was adopted after calculations showed that fitting an elliptical orbit to the observations made no significant improvement to their representation. The epoch is T. The coverage of the velocity-curve is good, but there are some large residuals. No M-K classification has been published: Griffin suggests that the star is a giant.

98. The new results are in good agreement with Petrie's earlier work (Pub. DAO, 7, 105, 1938) and in fair agreement with that of Ludendorff (ApJ, 25, 320, 1907). Coverage of periastron is incomplete because the period is so close to an integral number of days that periastron is unobservable from North America during this century. For this reason, Gorza and Heard used some of Petrie's observations to obtain their preliminary solution. The difference between Petrie's value of (24.2) and that found by Gorza and Heard may be real. If so, it represents an apsidal regress, perhaps to be ascribed to some form of periastron effect in this very eccentric orbit. A detailed abundance analysis of the spectrum has been published by J. Mitton (A&AS, 27, 35, 1977). The value of K is taken from the paper by J. Tomkin and H. Tran (AJ, 94, 1664, 1987), who also give e = 0.88.

99. Both spectra are visible. Batten and Szeidl thought the primary spectrum might be as late as A2. The secondary spectrum is similar but the K line in it is relatively weak, and the secondary might be an Am star.

100. The reference given in the Catalogue is to an abstract and most of the information given here is taken from a preprint of the full paper. The spectral type of the secondary is estimated from the effective temperature computed from the light-curve by Schiller and Milone. The epoch is the time of primary minimum and the orbit is assumed circular. Schiller and Milone do not compute K and K directly (the values given are estimates from their velocity-curves); they solve light-curves and velocity-curve together by the Wilson-Devinney method and find a mass-ratio of 0.6 and a total mass of 2.5 M. They believe that the primary nearly fills its Roche lobe. They find an orbital inclination of about 84 and a visual magnitude difference of 1.85 between the components. The star is a member of the cluster NGC 752.