To be a standard, or not to be?  _ Variable polarization
       in 9 Gem



       Masafumi Matsumura1, Munezo Seki2, and Koji Kawabata23

       1 Faculty of Education, Kagawa University, Takamatsu 760, Japan

       2 Astronomical Institute, Tohoku University, Sendai 980-77, Japan

       3 National Astronomical Observatory, Mitaka 181, Japan



       Abstract.   The linear polarimetry has been made in 1991-1996 for the
       supergiant star 9 Gem, which has been recently found to be periodic
       variable (period = 13.70 day) by the Hipparcos observation.  We find
       that the polarimetric variability marginally depends on the phase of the
       flux. This finding suggests that a portion of the observed polarization in 9
       Gem is circumstellar origin. This star will be still useful as a polarimetric
       standard of the position angle with the accuracy of 1 degree if the effect
       of periodic variability is calibrated.



1.   Introduction


A supergiant star 9 Gem (=HD43384, B3Ia) was one of the polarized standard
stars listed by Serkowski (1974), and had been used to calibrate the position
angles in linear polarimetry.  However, this star was claimed unsuitable for a
standard, because the temporal variability of polarization was reported (Hsu
and Breger 1982, Dolan and Tapia 1986). A serious problem by the discard of
this star is that we have no alternative bright polarized standards in the right
ascension about 6h.
     Recently, the catalogue by Hipparcos satellite has been released (ESA 1997).
In this catalogue, 9 Gem (=HIP29840) is shown to be periodically variable in
flux (the period is 13.70 day, and the epoch is J.D.=2448509.85). In this paper,
we investigate the dependence of polarimetric quantities on the phase of the flux
variation, to find a clue to the mechanism of the polarimetric variabilities, and
also to know whether 9 Gem is useful as a standard or not at a certain accuracy.



2.   Observations


We have made linear polarimetry of 9 Gem since 1991 with a multi-channel
polarimeter at Dodaira Observatory (Kikuchi 1988).  The stellar light into the
polarimeter passes the polarizer, then it is divided by dichroic mirrors, and fi-
nally it is guided to 8 photomultipliers. We thus obtain the polarimetric data in
8 bands simultaneously. The instrumental polarization was calibrated by observ-
ing the unpolarized standard stars (Serkowski 1974). The de-polarization factor
was obtained with the measurements of the stellar light through a Glan-Taylor
prism. The observations of 9 Gem were made with the monitoring observation
of R Mon (Matsumura, Seki and Kawabata, 1998).


    ( Figure 1.    The dependence of Pave on phase. )



3.   Results and Discussion


Fig.1 shows the polarization degree Pave of 9 Gem as a function of phase, where
Pave is the average over the data of 7 channels (0.36-0.76 m).  We have not
included the data of the 8th channel whose sensitivity is low. The polarization
degree Pave reaches its maximum at phase ~0.6, and decreases till phase ~0.4.
The amplitude is 0.2%.
     The difference of position angle between 9 Gem and that of HD21291 is
presented in Fig.2.  The star HD21291 is another polarized 'standard' listed
by Serkowski (1974), though it is noted that the position angle of HD21291
is slightly variable (Dolan and Tapia 1986). In this paper, we assume that the
position angle of HD21291 is constant. Then, Fig.2 shows
that the position angle of 9 Gem depends on phase.  The curve in Fig.2 is the
result of non-linear fitting on the assumption of a sin function. The amplitude
of the curve is 0.5 degrees.
     Those results suggest that at least a portion of the observed polarization is
stellar or circumstellar origin. If the star pulsates non-radially, dust might form
anisotropically in the circumstellar space.  This might cause anisotropic light
scattering, and might explain the observed variable polarization.


   ( Figure  2.    The  difference  of  position  angles  between  9  Gem  and
     HD21291 as a function of phase. )




4.   Conclusions


Our answer of the question in the title is as follows:

   o  Let 9 Gem be a standard, if the accuracy of the position angle may be as
      large as 3 degrees.

   o  If the accuracy of 1 degree is required, one should calibrate the periodicity
      of the position angle, to let 9 Gem be a standard.



References


Dolan, J. F. and Tapia, S. 1986, PASP, 98, 792

ESA 1997, The Hipparcos and Tycho Catalogues, ESA SP-1200

Hsu, J. and Breger, M. 1982, ApJ, 262, 732.

Kikuchi, S. 1988, Tokyo Astron. Bull. Ser. II, 281, 3267

Matsumura, M., Seki, M., and Kawabata, K. 1998, AJ submitted

Serkowski, K. 1974 in Planets, Stars and Nebulae Studied with Photopolarimetry,
       ed. T. Gehrels (University of Arizona, Tucson), p.135