Astron. Astrophys. 350, 491-496 (1999)

2. Stars with positive binary detection

The stars for which we could positively detect a companion are listed in Table 3, where the entries follow the style introduced in Paper I. In particular, the columns list the absolute value of the fitted angular rate of the event V in "/s, its deviation from the predicted rate as computed by us, the local lunar limb slope , the true position and contact angles, the signal-to-noise ratio (SNR), the projected separation, the brightness ratio and finally the K magnitudes of the two components in each star, based on the values given in Table 1. A more detailed explanation and discussion of these quantities is given in Paper II. In the following, we have tried to assign a spectral type to the companions, on the basis of the spectrum of the primaries and of the observed difference in magnitude. This was done on the basis of our observed K only. In principle, it should be possible to put better constrains on the spectral types by including m values at other wavelengths when available, but we note that in the case of previous speckle measurements this information is often missing, while in the case of previous LO measurements the adopted filters were often significantly different from the standard ones. Therefore, this task would require detailed computations using synthetic spectra and filter transmissions, which we did not attempt.

Table 3. Summary of binary detection results

2.1. SAO 162001

A companion has been detected for the first time around this star with a projected angular separation =36 milliarcseconds (mas), and K=3.4 mag. We constrain the spectral type of the companion to be an early F dwarf. This source was observed earlier in the visual by LO, but the companion was not detected (Edwards et al. 1980).

2.2. SAO 96746

SAO 96746 is a multiple system with a large angular separation of between components A and B. Component A was reported as an occultation binary by Dunham (1977), with =45 mas and PA=300^o. After this, several attempts by LO (Africano et al. 1978; Eitter & Beavers 1979) and speckle observations (Hartkopf & McAlister 1984), all at visual bands, failed to confirm the detection. Our earlier attempt by LO in the near-IR (Paper II) was also negative, and we speculated that the companion might have had a significant orbital motion in the time since Dunham's original detection. Now we are able to confirm the companion component A from the observations reported here. We find =14 mas along PA=120^o, and K=3.4 mag. By coincidence, our measurement occurred exactly as the same position angle of Dunham's occultation, although in the opposite direction: this confirms the fact that the companion must have an orbit which changes its position significantly over 20 years. Our previous negative detection can be explained by a combination of this effect, and the large K of the companion.

2.3. IRC -20444

We detect for the first time a companion around this late-type star, with =28 mas and K=2.8 mag. The spectral type of the companion can be constrained either to an early M giant or to an early O main sequence star. The stellar disk of the primary component was also resolved from our observations, and the results are reported in Richichi et al. (1998).

2.4. SAO 161202

This sub-arcsecond binary was resolved earlier by LO (Evans & Edwards 1981), and subsequently by speckle (Mason 1996). Our observations yield =178 mas along PA=94^o, and K=0.07 mag. The companion is likely to be of similar spectral type as that of the primary component.

2.5. SAO 162521

A companion has been detected around this star for the first time with =21 mas and K=3.1 mag. The spectral type of the companion can be constrained to around A5 dwarf. This source was repeatedly observed earlier by LO in the visual, but the companion was never detected (Africano et al. 1976; Eitter & Beavers 1979; Morbey et al. 1978). The angular diameter of the primary component was derived to be 6.50.5 mas, based on a model fit to the first diffraction fringe of a visual occultation lightcurve (Morbey et al. 1978). These authors also mentioned that a point source model fitted well the other fringes. Our observation rules out the possibility of such a large angular size for the primary.

2.6. SAO 94060

SAO 94060 was observed with negative results by earlier LO observations in the visual (Evans & Edwards 1981; Radick & Lien 1982), but was subsequently resolved by speckle (McAlister et al. 1987; McAlister et al. 1989). These latter authors report angular separations and position angles at several epochs. Our observation provides =567 mas along PA=59^o, and K=0.9 mag. We constrain the spectral type of the companion to be around G2, but we also note that using m from Hipparcos a type close to G8 is inferred.

2.7. AG +16 403

This star is a member of the Hyades cluster and is a spectroscopic binary with a period of 277 days, as reported by Griffin et al. (1985). A companion to AG +16 403 was also detected by speckle interferometry by Mason et al. (1993), who however recognized that the speckle component, at a separation of , is not the spectroscopic binary. In fact, they assign to the speckle binary a period of 167 years. Also our LO detected companion has a relatively large projected separation, =241 mas, and presumably it is not the spectroscopic but rather the speckle binary, also dubbed CHARA 154. However, after taking into account the position angles, we note that our result is not completely consistent with that obtained by speckle, at least under the assumption that the period is indeed much longer than the 6.2 years that separated the two measurements. We also note that in their original work, Griffin et al. (1985) mentioned that the analysis of the spectroscopic data for this system showed systematic anomalies that could not be well accounted for, and that their solution was only tentative. In the light of the more recent speckle and LO measurements, it might be worthy to investigate again the radial velocity data for this triple system. With a K=1.0 mag, the spectral type of the speckle/LO companion can be constrained to a late K dwarf.

2.8. SAO 98270

SAO 98270 is a speckle binary (Hartkopf et al. 1997), for which we derive a projected angular separation =177 mas and K=0.6 mag. The spectral type of the companion can be constrained to be around K5.

2.9. SAO 94554

SAO 94554 is a member of Pleiades group and a LO binary (Africano et al. 1978; Evans & Edwards 1981; Radick et al. 1982). Subsequently, the system has been studied extensively by speckle (McAlister & Hendry 1982; McAlister et al. 1989; McAlister et al. 1990; Mason 1996; Hartkopf et al. 1997; Fu et al. 1997). Recently, a binary orbit for this occultation binary has been published by Mason (1997), with a period of 15.38 yrs. The spectral type of the companion has been suggested by different authors to be between B8V and A0V. Our observation in the near-IR yields =15 mas, in good agreement with the orbit computed by Mason (1997), and a K=1.0 mag. This latter value leads us to assign a spectral type close to B9V for the companion. This is in agreement with previous estimates, but it leaves open the matter of the high combined mass for this system, as noted by Mason (1997).

2.10. SAO 93950

SAO 93950 was reported as a possible LO binary with a projected angular separation of 24 mas along PA=238^o and m=2.5 in a wide H filter by Fekel et al. (1980). However, there was no confirmation in spite of several other LO (Morbey et al. 1978; Evans & Edwards 1981; Radick & Lien 1982; Radick et al. 1982), as well speckle observations (Hartkopf & McAlister 1984; Mason 1996), all in the visual. We obtained two measurements of this source, the second of which confirms the companion. It is interesting to note that our positive detection occurred at a PA almost identical (see Table 3) to that of the event which lead to the discovery by Fekel and co-authors. This is not surprising, since the two observations are separated by almost exactly one Saros cycle, and the stations are at very similar latitudes. It is more intriguing the fact that we obtained also the same separation as in the original detection, but in the opposite direction. It would then appear that the companion has undergone a substantial orbital motion.

We note that the SNR value reported in Table 3 could give the impression of a detection below the noise level. In fact, this is a global value which incorporates the effect of atmospheric scintillation. The disappearance of the companion occurred after that of the primary, i.e. in a portion of the lightcurve where the SNR was close to 100. Reversely, the negative detection in our first observation, in spite of a formally higher SNR, can be explained by noting that one half of the first trace is affected by scintillation at a level which is larger than the magnitude of the companion. Also, our first and second events occurred along almost orthogonal position angles, and it is possible that the projected separation during the first event could have been too small for detection by our technique. The observed K=4.7 mag is consistent with a spectrum close to F2V for the companion.

Note also that the angular diameter of this star was derived in the visual and near-IR, with values of 1.631.07 mas and 2.270.24 mas respectively (White & Feierman 1987; Ridgway et al. 1982). Given the large brightness ratio, the influence of the companion on such measurements should be negligible. Also our measurements resolve the disc of the primary, and we will report our value for the angular diameter of this star elsewhere.

2.11. Remaining stars

There is no literature available for DO 2779, SAO 93083, SAO 93746, GCVS 980. Our lightcurves reveal a companion around these sources for the first time, with the parameters listed in Table 3. We note that SAO 93746 is best fitted by a triple star model, as shown in Fig. 1.

Fig. 1. Example of a new detection, the triple star SAO 93746. Left : Occultation data (dots), and best fit (solid line) by a single point source model. The residuals are shown in the lower panel. Center : Same, for a model with three point sources (see parameters in Table 3). The occultation times of the three components are marked. Right : Brightness profile, reconstructed by a model-independent method (see Paper II).

The last two stars, SAO 94002 and SAO 184176, are visual doubles with angular separations exceeding one arcsecond, and are reported here only for completeness. A rich amount of astrometric observations are available for SAO 94002 (Thé 1975; Jeffers & Vasilevskis 1978; Pannunzio & Morbidelli 1983; Jasinta et al. 1995). Both components of this well separated double system have been searched for multiplicity by speckle observations in the visual, with negative results (Mason 1996). There is no literature available on SAO 184176.

© European Southern Observatory (ESO) 1999

Online publication: October 4, 1999
helpdesk.link@springer.de