4 Comments on individual systems

ADS 455 = HR 134. The bright primary is a K0III giant which was considered as a spectroscopic binary in the Bright Star Catalogue (Hoffleit & Warren 1991 - BS). In fact its radial velocity is constant within $\pm 0.5$ km s^-1, as shown by our measurements, the old data and the equality with $\gamma$-velocity of Bab. The small relative motion of AB since its discovery in 1847 and equal radial velocities indicate that this is a physical system.

The measurements of a faint secondary at $8.5\hbox{$^{\prime\prime}$}$ from A were extremely difficult, being possible only on nights with good seeing. Although the velocity was seen to be variable already in 1994, only now are we able to derive an orbit. The faint secondary B was classified as K0V, which matches its color B-V=0.83. However, our orbit predicts the minimum masses of 0.87 and 0.84 $M_{\hbox{$\odot$}}$ implying somewhat earlier spectral types, G6V and G8V. Assuming these spectral types (which match also the dip EW), we run into another problem, because at the distance given by Hipparcos a pair of such stars must be $0.7^{\rm m}$ brighter than observed. A possible explanation might invoke inter- or circumstellar absorption $A_V=0.7^{\rm m}$, which would also account for the redder color. Alternatively, the errors in parallax and photometry may be at the origin of this discrepancy. Eclipses may be searched in this system, because the inclination i is close to $90\hbox{$^\circ$}$. Axial rotation is synchronized with orbit, since the expected equatorial velocity 6.7 km s^-1 is close to the actually measured $V\sin i$.A 7-day dwarf binary is expected to have circular orbit; small but significant eccentricity of ADS 455B can be explained by the perturbations from the visual component A (Mazeh 1990).

ADS 497. We discovered the close sub-system independently, realizing later that an identical single-lined orbit has already been published by Latham et al. (1988). However, we also measure the secondary dip, and publish here a double-lined orbit which incorporates the data of Latham et al. These two data sets are comparable in quality (rms unit-weight residuals for the primary are 0.53 and 0.65 km s^-1 for the 23 and 25 observations of RVM and CfA spectrometers, respectively), with no significant difference of zero points. This triple system belongs to the Arcturus moving group and may be about 10¹⁰ yrs old (Eggen 1971). Hipparcos astrometry and new radial velocities only confirm this conclusion, giving the galactic velocity U, V, W = 18, -113, -56 km s^-1. The components are located on the Main Sequence close to the turning point of the oldest clusters, and in few billion years the primary will start to expand. Our analysis of dip contrast gives metallicity ${\rm [Fe/H]}=-0.6$ for both components, whereas Latham et al. (1988) give ${\rm [Fe/H]}=-1.16$.

  

Table 5: Models of multiple systems

ADS 1134. This object was placed on the observing program only in 1998, and the double-lined nature of C was discovered immediately. The velocity amplitudes and dip parameters of Ca and Cb are indistinguishable. Minimum masses are close to those estimated from spectral types, and eclipses are hence possible ($i \approx 90\hbox{$^\circ$}$), although Hipparcos did not detect any brightness variations. Both for A and C the EW of dip indicate solar metallicity. The spectral type of the visual secondary B given in Table 5 was assigned to match the magnitude difference. In summary, ADS 1134 presents a quadruple system composed of solar-type dwarf stars with perfectly normal masses and luminosities.

ADS 5436 = HR 2486/85. The visual components are so close on the sky and so similar in brightness that they were frequently confused. As shown in the Hipparcos catalogue, the western component, designated as B in ADS, is slightly brighter, and it is HR $2486={\rm HD}~48766$. It is this star that is found here to be a single-lined binary. Some of our measurements were influenced by the light from A entering into the slit, which explains the few deviating points in Fig. 1. Undoubtedly, this problem was common to other observers, and, together with frequent misidentifications, could be the reason why the radial velocity of A was also considered as variable. In fact it is constant, and the difference of 2.2 km s^-1 with the $\gamma$-velocity of B can be accounted for by the orbital motion in the wide pair AB. Since the discovery of AB in 1830 the separation decreased from $4.8\hbox{$^{\prime\prime}$}$ to $4.6\hbox{$^{\prime\prime}$}$without change in position angle, indicating either a very eccentric or a highly inclined orbit. The component C is definitely optical (cf. Table 2), as confirmed by our radial velocity measurement. The EW of A and B are compatible with solar metallicity. If Ba rotates synchronously with orbit, its equatorial velocity must be 14.1 km s^-1, to be compared with the measured $V \sin i = 17.4$ km s^-1.

ADS 8236. This pair of nearby dwarfs was placed on our observing list in 1988 on request of Dr. A.A. Kiselev in order to measure accurately the relative radial velocities of the components. The slow variations of the radial velocity of B were discovered in 1990, and now the two periastron passages of this eccentric 4.6 yr orbit are covered by observations. After calculation of the first preliminary orbital solution it was realized that the secondary must be relatively massive. Indeed, in 1997 we obtained 3 observations of its dip, which has only 2% contrast. This fixes the secondary mass and enables to estimate the orbital inclination, which is close to 90$^\circ$. This system is similar to ADS 9167, for which a 2.87 yr spectroscopic orbit has been published by Kiyaeva et al. (1998). In both cases the semimajor axis of the spectroscopic subsystems is large, and the corresponding perturbations could be (but have not been) detected from the precise relative astrometry of the wide pairs. For ADS 8236 the semimajor axis of Bab orbit is 73 mas, and the semimajor axis of Ba motion around the center of gravity is 30 mas. It seems likely that by reprocessing the raw Hipparcos data an astrometric orbit of Bab can be extracted. The Hipparcos double star solutions published so far did not consider the model of visual+astrometric triple systems, and this is why no such systems are found in the catalogue.

ADS 9444. This system, placed on the program by A.A. Kiselev, is in fact optical, as can be inferred from its fast relative motion. Radial velocities only confirm this assertion. The A component turned out to be a spectroscopic binary with a 8.9 yr period. The minimum mass of Ab is around 0.5 $M_{\hbox{$\odot$}}$. We do not give system model of ADS 9444 in Table 5. Component B has a dip of high contrast and may be a background giant.

ADS 10044. Also from Kiselev's list, this is actually a physical triple star. Hipparcos parallax places both components well above the Main Sequence: A and B are located at the very beginning of the giant branch of the HR diagram of NGC 188, as given in Fig. 2 of Eggen (1971). Comparison with evolutionary tracks (Schaller et al. 1992) indicates a mass of A around 1.1 $M_{\hbox{$\odot$}}$ and an age between 5 and 10 billion years. The object belongs to the old disk population, with spatial velocity of U, V, W = -39, -61, +2 km s^-1. Recently a triple system of similar evolutionary status, HD 158209, was discussed by Griffin (1997). The orbital period of its close pair, $21.825^{\rm d}$, is strikingly similar to that of ADS 10044A, $21.726^{\rm d}$. Both orbits have small eccentricities; it is possible that they have been partially circularized due to the evolutionary expansion of the primary components. Both these triple systems are at a distance of 100 pc, but they have distinctly different spatial velocities and hence do not belong to the same moving group.

ADS 11163. The only available information on this visual triple star, which is found to contain a short-period subsystem, comes from WDS (Worley & Douglass 1984). The relative position of AC ($\rho = 9.2\hbox{$^{\prime\prime}$}$)is fixed during 60 years, which, coupled with the proper motion of 60 mas/yr, indicates that C is physical component. The same reasoning applies to the closer pair AB ($\rho = 1.5\hbox{$^{\prime\prime}$}$). Taking the spectral type of the primary to be G5V (with such a short orbital period it can not be a giant), we ascribe the spectral types to B and C in order to match their magnitudes (cf. Table 5).

The light of A and B was not separated on the slit, but we believe that it is the A component which is variable. Our model predicts that the relative EW of the A and B dips are 64% and 36%, respectively, if their light is fully mixed. The velocity amplitude of the combined dip is reduced in the same proportion. The actual amplitudes would be 6.7 or 11.9 km s^-1, depending on whether it is A or B which is variable. These amplitudes are comparable to the dip FWHM of 14.5 km s^-1, and variations of dip parameters with orbital phase are to be expected. In the plots of the contrast and width of the dip against the orbital phase the variations are barely detectable, being of the order of 10%. So, we conclude that the short-period subsystem belongs to A. Corrected velocity amplitude of 6.7 km s^-1 corresponds to the minimum secondary mass of 0.05 $M_{\hbox{$\odot$}}$. The inclination is not known, but this system is candidate for a substellar mass object in a short-period orbit; no such objects have been found so far.