COMMISSION 27 OF THE I. A. U.
           INFORMATION BULLETIN ON VARIABLE STARS
                       Number 2881

                                        Konkoly Observatory
                                        Budapest
                                        17 March 1986

                                        HU ISSN 0374 - 0676



              THE Mg II LIGHT CURVE OF AR Lac



  AR Lac (HD210 334) is one of the most important binary systems of the
RS CVn type. Numerous observations have been performed in the ultraviolet
wavelength (1100 A - 3200 A) by the IUE satellite. Presently we are working on
the long wavelength images of AR Lac released from the IUE archive. Preliminary
results of a limited number of observations have been reported previously 
(Kiziloglu, 1983). We want to thank all astronomers who made the observations 
and also the IUE project team who made the archive data available to us.
In this work, 23 high resolution and 16 low resolution long wavelength (2000 A 
- 3000 A) images are used. The images cover the observation period of about
five years between 1979 and 1984.
   In Figure 1b, variation of fluxes obtained from the Mg II h+k resonance
lines as a function of the orbital phase is shown. During the calculation of
the fluxes, the main problem was the systematic difference of the flux values
between the high and low resolution spectra. To our knowledge, no clear
calibration analysis has been given in the literature for this discrepancy.
To remove this systematic difference, all the high resolution Mg II spectra
were degraded using the 8 A Gaussian instrumental profile. The Mg II fluxes
calculated from the degraded spectra are then compared to the fluxes calculated
from the low resolution Mg II fluxes. For this comparison, images with the
observation times close to each other are considered only to avoid a possible
time variability of the activity. This process indicated that the low resolution
fluxes were smaller than the high resolution Mg II h+k fluxes by a factor
of 1.6. Thus all the low resolution fluxes multiplied by this correction
factor are plotted accordingly in Figure 1b.
  A close inspection of the figure after this correction reveals the 
following:
  i) The depths of the eclipse features of the Mg II light curve at the
phases 0.0 and 0.5 are almost the same. The flux values at mid eclipses are
about 70 % of that at the quadratures. This suggests that the Mg II emission
from both components is the same for unit area on their surfaces.


                            [FIGURE 1]

 Figure 1. (a) Visual light curve of AR Lac (Ertan et al., 1982);
 (b) The Mg II h+k light curve of AR Lac between 1979 and 1984. Filled
 circles represent flux values observed in a single orbit (5-7 October
 1983). Phases were calculated using Min I = J.D.Hel.2443672.0917 +
 1.98318235d E.


   ii) Eclipse features do not show large flux variations globally during
the observation period of about 5 years indicating the presence of quite
stable chromospheres of both components.
   iii) The eclipse profile is highly symmetrical at the phase 0.5 whereas
asymmetry is observed at the phase 0.0. The fluxes are scattered around the
latter phase. A noticeable variation is also observed at the quadratures.
The reason for this variation can be the change of the chromospheric activity
within the five years of observations or short time scale activities during
the observations. In the case of asymmetry several suggestions can be made;
however, the small number of the observations prevents further comments.
   iv) Comparing the V light curve (Figure 1a, Ertan et al., 1982) and the
Mg II h+k light curve (Figure 1b) we notice that the eclipse durations are
longer for the Mg II light curve. This points out that each companion has an
extended chromosphere. Assuming that only the photosphere of one component
can eclipse the chromosphere of the other, thicknesses of the chromospheres
can be estimated geometrically. Results of such an analysis including all



observational points show that the chromospheric radii of the G and K stars
of the AR Lac are 4.2 +- 1.0 and 5.7 +- 1.1 Rsun respectively. The photospheric
radii for the G and K stars are 1.54 and 2.81 Rsun respectively (Chambliss,
1976). On the other hand, if the same procedure is applied to the points
observed on a single orbit (day 278 in 1983, filled circles), 3.2 +- 0.4 and
4.6 +- 0.3 Rsun are obtained as the chromospheric radii for the G and K stars
respectively. These values show that the chromospheres possibly extend to the
first Lagrange point of the binary system. The above values agree with the
values given by Naftilan and Drake (1977) but disagree with those given by
Walter et al. (1983). Since the radii of the chromospheres of the binary
components of the AR Lac calculated in this work are based on the observational
fluxes within the accuracy of the IUE satellite, these values should be 
considered more reliable.

   Detailed results will be published elsewhere.




                            E. DERMAN

                       Department of Astronomy and Space
                       Sciences, University of Ankara
                           Ankara, Turkey


                       U. KIZILOGLU and F. TOKDEMIR

                          Physics Department
                       Middle East Technical University
                         (ODTU), Ankara, Turkey


References:

Chambliss, C.R., 1976, Publ. Astron. Soc. Pacific, 88, 762. [BIBCODE 1976PASP...88..762C ]
Ertan, A.Y., Tumer, O., Tunca, Z., Ibanoglu, C., Kurutac, M., and Evren, S.,
    1982, Astrophys. Space Sci., 87, 255. [BIBCODE 1982Ap&SS..87..255E ]
Kiziloglu, U., Derman, E., Ogelman, H., and Tokdemir, F., 1983, Astron.
    Astrophys., 123, 17. [BIBCODE 1983A&A...123...17K ]
Naftilan, S.A., and Drake, S.A., 1977, Astrophys. J., 216, 508. [BIBCODE 1977ApJ...216..508N ]
Walter, F.M., Gibson, D.M., and Basri, G.S., 1983, Astrophys. J., 267, 665. [BIBCODE 1983ApJ...267..665W ]