COMMISSION 27 OF THE I. A. U.
         INFORMATION BULLETIN ON VARIABLE STARS
                      Number 2060

      Rothney Astrophysical Observatory Publication
                    Series B. No. 8


                                Konkoly Observatory
                                Budapest
                                1981 December 17

                                HU ISSN 0374-0676




  OPTICAL OBSERVATIONS OF THE PRIMARY MINIMUM OF THE SOLAR-
        TYPE BINARY AI Phe WITH THE IUE SATELLITE



    In an extension of a program to study the ultraviolet center-to-limb
variation in the sun, we have used the International Ultraviolet Explorer
Satellite (IUE) to observe the flux during the eclipse of the G2 component of
the 24.6d totally eclipsing system AI Phoenicis.
    Previous uvby studies of this system reported by R. Reipurth (IBVS No.
1419, 1978) established that the primary eclipse was total. M. Imbert's
(Astron. and Astrophys. Suppl. 36, 453, 1979) spectroscopic study indicated
that the apparently cooler component was of spectral type G5, slightly more
massive, and slightly more luminous than the hotter star.
    We anticipated that, although the cooler star might be slightly evolved,
the components would not be strongly interacting and that the conditions of a
lengthy eclipse (13h) and relatively bright system magnitude (8.5 <= V <= 9.2)
would permit a sufficient number of IUE exposures to determine the UV limb-
darkening.
     On UT August 20-21, ten LWR spectra were obtained in two successive
eight-hour shifts at Goddard SFC and at Vilspa, covering totality and the
rising branch. Features are discernible to at least 2600A on the least well
exposed frame. Fine Error Sensor (FES) flux readings were taken before and
after each exposure to tie in the light curve variation. An algorithm of
A. Holm and G. Rice (private communication, 1981) was used to convert FES
flux to V magnitude:


    V = -2.5 log(C/(1-1.6d * ex(-4) * C^0.781)) + 16.58 - 0.24 (B-V).
            



The results appear in Figure 1, with the time of IUE exposures indicated by
shaded bars. The phase is computed using Reipurth's ephemeris, with which
our data appear to be in reasonable agreement.

                     [FIGURE 1]

    One of us (BJH) obtained ground-based UBVRI observations and spectra
at CTIO and another of us (I. Shelton, resident observer of the David
Dunlap Observatory, University of Toronto Station at Las Campanas) obtained
UBVRI observations at Las Campanas. The value of B-V used in the above
equation came from a preliminary reduction of these ground-based data. The
reduced light curves will be modeled in five colors, and the resulting
improved elements used in the determination of the center-to-limb variation
in the eclipsed component. The V data from the different sources show
rough agreement in amplitude and shape except at the high shoulder. These
results suggest that the duration of totality is at least 105 minutes and
perhaps as long as 140 minutes, therefore longer than the 80 minutes
of totality in Reipurth's report. The fully reduced photometric data and
the spectra will be discussed elsewhere.





        E.F. MILONE, B.J. HRIVNAK, T.A. CLARK,
        RAO, University of Calgary,

        O. KJELDSETH MOE,
        Institute of theoretical Astrophysics,
        University of Oslo,

        J.C. BLADES, Resident Astronomer,
        Villafranca del Castillo Satellite Tracking
        Station, Madrid,

        I. SHELTON,
        DDO, The University of Toronto.