COMMISSION 27 OF THE I. A. U.
         INFORMATION BULLETIN ON VARIABLE STARS
                    Number 2201

                                 Konkoly Observatory
                                 Budapest
                                 1982 September 29

                                 HU ISSN 0374-0676





    PERIODIC VARIATIONS IN THE POLARIZATION OF V 603 Aql +





In August and September 1981 the linear and circular polarization 
of the old nova V 603 Aql has been observed (partly simultaneously 
with photometric observations) using the ESO two channel 
polarimeter at the 1m and 61cm telescopes. By technical
reasons the observations have been performed during two sections
separated by a gap of 11 nights. From the radial velocity
measurements published by Kraft (1964) a low inclination angle
of the nova system must be deduced. Therefore an extremely weak
intrinsic polarization has to be expected and only a large number 
of measurements allows to draw conclusions about the system.
This conflicts with the long integration time necessary for a
single polarization measurement of a V approx. 11.9 star (Haefner,
1981). As a compromise all measurements have been carried out
with an integration time of 160 sec in the white light. 
Altogether 470 linear and 310 circular observations have been
collected. The instrumental polarization was determined each
night by measuring near distant stars. Since there is a clear
difference between the spectroscopic period (3h19.5m) reported
by Kraft (1964) and the photometric period (3h28.6m) found by
Haefner (1981) it was a main point to check the polarimetric
results for periodicities by correlation and periodogram 
techniques.





+ Based on observations collected at the European Southern
 Observatory.




In the linear polarization no high significant period could be
detected in the range from 3 min. to 6 hours. Only one period
with 3h33.8m was more pronounced (see Fig. 1). Since this
period does not coincide either with the spectroscopic or the
photometric one an interpretation is difficult at the moment.
However, it should be emphasized that this periodic variation
was completely absent during the first section of observation
and may be correlated with a special hump phenomenon in the
light curve: During the second observation section an increase
in the intensity with a duration of approx. 3 min. and an amplitude
of approx. 0.2m could be observed sporadically. This peak was 
superimposed to the broad hump observed by Haefner (1981) and had a
mean period of about 3h39m. Shortly before and after this peak
a remarkable increase in the scatter of the polarization values
could be noticed.

                       [FIGURE 1]

 Fig. 1: Phase diagram of the degree of the linear polarization
         during section 2 (315 values). The vertical bars give
         the mean error within phase bins of 0.2.


In the circular polarization a highly significant period of
3h18.2m is present (Fig. 2). In terms of the usual model of


                       [FIGURE 2]

 Fig. 2: Phase diagram of the normalized Stokes parameter V (310
         values). The vertical bars give the mean error within 
         phase bins of 0.2.


cataclysmic variables this polarization has to be ascribed to
the white dwarf. Shape and change in sign of the Stokes 
parameter V suggests a magnetic rotator. However this would imply
that the white dwarf would rotate almost synchroneously like
the AM Herculis type stars.



More detailed results which also take into account 
spectroscopic and photometric measurements will be published 
elsewhere.





               K. METZ

               Universitats-Sternwarte
               Scheinerstr. 1
               D-8000 Munchen 80
               Germany





References:
Kraft, R. P. 1964, Astrophys. Jour. 139, 457 [BIBCODE 1964ApJ...139..457K ]
Haefner, R. 1981, IBVS, No. 2045