COMMISSION 27 OF THE I. A. U.
        INFORMATION BULLETIN ON VARIABLE STARS
                    Number 2013

                                 Konkoly Observatory
                                 Budapest
                                 1981 September 7

                                 HU ISSN 0374-0676





       5 Ceti = HR 14 : A NEW AND PUZZLING VARIABLE STAR


  We report photoelectric photometry which shows 5 Ceti to be a variable
star with an amplitude a bit over 0.2m in V and a period apparently very
close to the orbital period of 96.41d determined spectroscopically by 
Christie (1933). According to the 1964 edition of the Yale Bright Star 
Catalogue the spectral type is gK2.
  Altogether this star was observed on 28 nights in 1979 (between JD
2444138.8 and 2444222.6) and 46 nights in 1980 (between JD 2444445.9 and
2444563.7) using 29 Piscium as the comparison star. The telescope was a
20-inch Cassegrain reflector at the Lines Observatory near Mayer, Arizona
equipped with an unrefrigerated 1P21 photomultiplier operated at -900 V.
   Nightly means are plotted in the figure, where Delta v_0 is differential
magnitude in the sense variable minus comparison, corrected for differential 
atmospheric extinction but not yet transformed to V of the UBV system.
Phase in this figure is computed with the ephemeris

     JD(hel.) = 2444176.5 + 96.41d n,
where the period is the orbital period of Christie (1933) and the initial
epoch is a time of primary minimum derived from this photometry. If we
consider all of the relevant uncertainties, it appears primary minima are
occurring at times of conjunction (with the gK2 star behind).
   The fact that 5 Ceti is a spectroscopic binary, the shape of the light
curve, and the occurence of minimum light at times of conjunction all suggest 
that 5 Ceti is an eclipsing binary. The puzzle is that a binary with
such a long period can produce a light curve with a W UMa shape. Presumably 
one (or both) of the two stars is large enough to fill or nearly fill
its Roche lobe; this would be so if one (or both) of the stars is as large
as ~50 R_Sun. Perhaps we should not be so surprised, since the one star
seen in the spectrum has been classified a giant. Because that gK2 star
is behind at primary minimum, we can say that the other star is cooler
(spectroscopically later) than K2.

                           [FIGURE 1]

   For 29 Piscium the 1964 Yale Bright Star Catalogue gives V = 5.10m.
From this we see that 5 Ceti varies between V = 6.1m and 6.3m. Spectroscopists
should look for the fainter star in the spectrum, and photometrists should
get the complete light curve in other bandpasses, such as B and U and perhaps 
R. In future photometry a better comparison star might be 4 Ceti.
Though much bluer than 5 Ceti (29 Piscium was also), it is much closer in
the sky and more nearly the same brightness.


                       RICHARD D. LINES

                       6030 North 17th Place
                       Phoenix, Arizona 85016
                       U.S.A.

                       DOUGLAS S. HALL
                       Dyer Observatory
                       Vanderbilt University
                       Nashville, Tennessee 37235
                       U.S.A.



Reference:
Christie, W. H. 1933, Ap.J. 77, 310. [BIBCODE 1933ApJ....77..310C ]