COMMISSION 27 AND 42 OF THE IAU
                   INFORMATION BULLETIN ON VARIABLE STARS
                               Number 3809

                                                         Konkoly Observatory
                                                         Budapest
                                                         20 November 1992

                                                         HU ISSN 0374 - 0676


     A Detection of Rapid Infrared Oscillations in HR 1217=DO Eri

The rapidly oscillating Ap (roAp) star HR 1217 was observed through a Johnson K
filter for 2.4 hr on the night 17/18 August 1992 (JD 2448852) with the SAAO
1.9-m Radcliffe telescope and Mk III InSb photometer. The f/50 secondary mirror
chops between two apertures to measure the quantities (star+sky1)-sky2 and
sky1-(star+sky2). Subtraction of these quantities yields twice the stellar
signal minus the sky background contribution. Since we were searching for rapid
oscillations with periods around 6 min., no standard stars were observed and we
did not attempt to put the observations on the standard system. The resulting
instrumental magnitudes were corrected for mean extinction.
   Figure 1 is the light curve of these data which shows that there was a
gradual improvement in sky transparency during this run. Fig. 2 shows the
Fourier transform of those data up to the Nyquist frequency of 6.0 mHz. The
peaks at low frequency are caused by sky transparency variations. Normally we
would remove these frequencies, but for illustrative purposes we show them in
Fig. 2 to convince the reader that we have a credible detection of oscillations
at the known optical oscillation frequency even in the presence of the sky
transparency variations.


                               [FIGURE 1]


   The peak labeled 'Nu_1' lies at 2.73 mHz, exactly the frequency of the
well-studied B-band oscillations (Kurtz et al. 1990). Removing the
low-frequency noise lowers the general level of the peaks but does not
otherwise change the appearance of the spectrum much. When the 4 highest peaks
in the frequency range 0-0.92 mHz are removed the highest peak remaining is at
2.72 mHz. A least squares fit of Nu_1=2.72 mHz to the prewhitened data yields
an amplitude of 13±0.4 mmag for Nu_1.


                             [FIGURE 2]


   Without prior knowledge of the optical oscillations at 2.72 mHz, this peak
would be statistically insignificant. However, the chance of finding a peak
of z signal/noise is power at exactly the right frequency is ~e^-z (Scargle
1982). Taking A_1 = 1.3±0.4 mmag and the level of the noise in the amplitude
spectrum of the prewhitened data as a conservative 0.8 mmag, z=(1.3/0.8)^2
which yields a False Alarm probability of 7% for this peak; i.e. the peak Nu_1
is significant at the 93% level.
   Unfortunately, the beating of the principal oscillation modes in HR 1217
does not allow us to predict the instantaneous B amplitudes on any given night.
Thus we cannot compare our tentative measurement of A_K to any expectations of
the K amplitude. For this, simultaneous multicolour photometry will be
necessary.


                             Peter Martinez  1, & K. Sekiguchi  2,


1, Dept. of Astronomy, University of Cape Town, Rondebosch 7700, South Africa
2, S.A. Astronomical Observatory, P.O. Box 9, Observatory 7935, South Africa


PETER @ UCTVAX.UCT.AC.ZA



References:

Kurtz, D.W. et al., 1989, Mon. Not. R. astr. Soc., 240, 881. [BIBCODE 1989MNRAS.240..881K ]
Scargle, J.D., 1982, Astrophys. J., 263, 835. [BIBCODE 1982ApJ...263..835S ]