5 Conclusion

We have shown that the cool Ap star HD 8441 belongs to a triple system, whose ratio of the long to the short period is larger than 47.

We could improve the knowledge of the orbit of the classical Ap star $\beta$ CrB thanks to our new, highly homogeneous radial velocities which could be combined with the published speckle orbit. Furthermore, the individual masses of the components could be determined thanks to the Hipparcos parallax.

The masses of both components of the Am star HD 43478 could be obtained with 1% accuracy thanks to the eclipsing nature of the system. The radii are less precisely known, because of the shallowness of the eclipses and the insufficient photometric data. Nevertheless, the assumption of synchronism combined with the present data leads to a very good match of both components with the isochrone $\log t = 9.1$,and the evolved state of at least the primary clearly appears even without this assumption. It would be worthwhile to make further photometric observations of this system, in order to obtain a better estimate of the radii and to tell whether it is synchronized or not. Spectroscopic observations should also be done in order to constrain the luminosity ratio and the individual effective temperatures.

Finally, we obtained a good mass ratio for the Am star HD 96391 and could also estimate its orbital inclination from an independant mass determination using the Hipparcos parallax.

All four binaries examined here have at least one component (the primary) which is significantly evolved and will leave the Main Sequence within a short time, relatively to its MS lifetime. The rotational period of both components is very probably synchronised with the orbital period in the two systems hosting Am stars.

Acknowledgements

This work was supported in part by the Swiss National Foundation for Scientific Research. Part of the reduction of the data was made by the late Dr. Antoine Duquennoy. The photometric data were reduced by Mr. Bernard Pernier, Mr. Christian Richard, Prof. Frédy Rufener and Prof. Gilbert Burki. We thank Dr. S. Hubrig for drawing our attention to Tokovinin's 1985 paper and 1997 catalogue.