Astron. Astrophys. 324, 97-108 (1997)

6. Conclusions

A detailed spectroscopic study of (  K, , ) has shown that it is a low-metallicity star ([Fe/H] ) with overabundances in the s-process elements. It is a twin of AG Dra studied in Paper I, another yellow symbiotic exhibiting the `barium syndrome'. The s-process abundance distribution is best fitted by a single neutron exposure of mb^-1, and a neutron density , as derived from the observed Rb/Zr ratio. These results confirm that the s-process synthesis is more efficient (larger and ) at lower metallicities, a result already inferred from the study of barium and CH stars. That trend had been predicted by Clayton (1988) in the case that is the neutron source responsible for the operation of the s-process.

The barium syndrome seems to be shared by many yellow symbiotics (, AG Dra, UKS-Ce1 and S32; see Schmid 1994), which also appear to be halo objects. This correlation may be a consequence of the fact that the barium syndrome is more easily produced at low metallicities. It is moreover argued that low-metallicity giants reach higher luminosities at a given than giants of lower metallicities, and thus suffer from a more severe mass loss making them appear as symbiotic stars.

The orbital parameters of have been derived from CORAVEL measurements, and are similar to those of barium stars. Like the other two yellow symbiotics with known periods (AG Dra and LT Del), the orbital period (282 d) of falls in the short-period tail of the distribution of barium stars. Periods for more yellow symbiotic stars are however required before one can conclude that this is a systematic effect.

The fact that exhibits ellipsoidal light variations has been used to narrow down the range of admissible parameters. However, no solution matching the Geneva evolutionary tracks could be found. This discrepancy has been interpreted as an indication that the usual Roche geometry is not applicable to systems where a strong mass loss from one component is driven by a force substantially reducing its effective gravity. This effect is also present in X-ray binaries, and makes the restricted three-body equipotentials around the (more massive) mass-losing component shrink. A solution matching the orbital parameters, the ellipsoidal lightcurve and the evolutionary constraints may be obtained when this effect is taken into account.

© European Southern Observatory (ESO) 1997

Online publication: May 26, 1998

helpdesk.link@springer.de