COMMISSION 27 OF THE I.A.U.
                 INFORMATION BULLETIN ON VARIABLE STARS
                             Number 3297

                                                         Konkoly Observatory
                                                         Budapest
                                                         1 March 1989

                                                         HU ISSN 0374 - 0676


                   On the lightcurve of HD27563

The B5III (Cowley, 1972) star HD27563 (= HR1363) has been studied photelectrically 
by Mathys et al. (1986) in the uvby system. They concluded that it was
probably a multi-periodic object with periods around 4 days. This was
essentially based on observations we made during 20 nights in November 1977 (43
measurements) at the ESO observatory on La Silla, with the 50cm telescope. A
few additional data came from two other observing runs we carried out in the
same system, at the same telescope, in December 1978 (7 measurements) and
September 1981 (7 measurements).
   More data have been provided by the "Long Term Photometry of Variable" group
(Sterken, 1983) during two months of observing at La Silla in October and
December 1982 (27 measurements). They are also in the Stromgren system,
although the filter set used at that occasion was less than perfectly conform.
   In all four observing runs, HD27563 was used as a comparison star for the
variable CP star HD29009. Coincidentally the period of that CP star is also
close to four days so that it was not immediately obvious that each object did
vary. The other comparison star which was measured proved to be variable too,
so that we had to rely exclusively on "absolute" reductions to study the
lightcurve of HD27563.
   Among period searching methods, Deeming's (1975) method is well suited to
multiperiodic phenomena. The power spectra obtained for the largest data sets
(1977 and 1982) show a clear peak close to nu = .25 d^-1 (Figures 1 and 2) as
well as strong aliases at 1 +/- nu.  However the central frequencies do not
coincide exactly (0.257 d^-1 in 1977, and 0.275 d^-1 in 1982). Because of the
longer time base, the 1982 data set is able to show separate components in the
peaks, and particularly at 0.254 d^-1, close to the 1977 value. The frequency
splitting (Delta nu ~0.021 d^-1) is only 20% wider than the natural splitting
shown in the spectral window, indicating some degree of interference. Power
peaks appearing around 1.0 d^-1 are aliases of very low frequency features-not
seen entirely on the figures-and are artefacts due to an improper normalization
between the data of October and December 82.
   After merging all data sets, Deeming's frequency analysis gives a very noisy
spectrum with a complex substructure (Figure 3). The strongest peak occurs at
nu_1 = 0.25503 d^-1 (or 3.9211 d) with fine structure components (barely
resolved on Figure 3) around nu_1 +/- 0.0022 d^-1. The broader splitting seen in
the 1982 data shows itself with main components around nu_2 = 0.27502 d^-1
(3.6361 d) and around nu_2 +/- 0.0022 d^-1. Strong aliases are present close to
1 +/- nu_1 and 1 +/- nu_2 and so are the combs around multiples of .5 d^-1.


                                [FIGURE 1]

 Figure 1: Power spectrum of the observations of HD27563 obtained in November
 1977. All colours show very similar behaviours. The geometric mean of u, 
 v, b and y is presented here. Units are mmag^2.

                                [FIGURE 2]

 Figure 2: Power spectrum of the observations of HD27563 obtained in
 October-December 1982.

                                [FIGURE 3]

 Figure 3: Power spectrum of all observations of HD27563.


   After prewhitening by nu_1 or any of its main secondary components, the
peaks around nu_2 subsist although with slightly different frequencies and
amplitudes. At the same time the 1 +/- nu aliases becomes more important.
Prewhitening of the data with several of those frequencies is always less than
satisfactory, suggesting the presence of additional frequencies. The main
frequency at nu_1 = 0.25503 d^-1 or 0.25284 d^-1 is apparently stable over
several years. Our analysis favours the presence of at least another frequency
nu_2, with the most likely pairings (nu_1, nu_2) being either of the following
combinations: (0.25503, 0.27785), (0.25503, 0.29403), (0.25284, 0.27723),
(0.25284, 0.27504). Because of the complexity of the power spectrum, it is
difficult to decide conclusively on the presence or stability of the
periodicities. However we notice that a quadruplet of frequencies (for
instance 0.228734, 0.255027, 0.277850, 0.289394) yields a very clean
prewhitening, and we suggest that such a multiperiodicity is effectively
present.
   The relatively bad accuracy of the data due to the lack of comparison stars,
as well as the closeness of the periods to 4 days, make further analysis of
HD27563 desirable.


                                [FIGURE 4]

 Figure 4: Power spectrum of the observations of HD27563 prewhitened for
 nu_1 = 0.25503 d^-1.


   Acknowledgments: The author is grateful to Dr. P. Renson for helpful
comments. He also thanks Drs. O. Stahl and M. De Groot for their contribution
as observers of the "Long Term Photometry of Variables" project.


                 J. MANFROID

             Institut d'Astrophysique
               Universite de Liege
                 B-4200 Liege
                   Belgium



References:

COWLEY, A.: 1972, Astron. J. 77, 750. [BIBCODE 1972AJ.....77..750C ]
DEEMING, D.J.: 1975, Astrophys. Space Sc. 36, 137. [BIBCODE 1975Ap&SS..36..137D ]
MATHYS, G., MANFROID, J., RENSON, P.:  1986, Astron. Astrophys. Suppl.
   Ser. 63, 403. [BIBCODE 1986A&AS...63..403M ]
STERKEN, C.: 1983, The Messenger 33, 10. [BIBCODE 1983Msngr..33...10S ]