COMMISSION 27 OF THE I.A.U.
                 INFORMATION BULLETIN ON VARIABLE STARS
                              Number 2932

                                                         Konkoly Observatory
                                                         Budapest
                                                         10 September 1986

                                                         HU ISSN 0374 - 0676


      B AND V LIGHT CURVES OF THE MASSIVE CLOSE BINARY DH CEPHEI

DH Cep (=HD 215835; SAO 34810; <V> = +8.58) is a double-lined, massive
spectroscopic binary consisting of a pair of O5-6 V stars having an orbital
period of 2.111 days. The spectrographic radial velocity study of Pearce
(1949) indicates an orbital eccentricity of e = 0.13 and values of M_1sin^3i=
23.4 M_sun and M_2sin^3i=19.1 M_sun. Photoelectric photometry of DH Cep was
carried out by Hill, Hilditch and Pfannenschmidt (1976) in which small (~0.05
mag) light variations were observed but the data were too few to define a light
curve. In agreement with this photometry, observations of the star in the
ultraviolet (lambdalambda1550-3300) made with the ANS satellite during 1975
and 1976 show only low amplitude brightness changes which appear to arise from
the distorted figures of the binary components (Wu and Eaton 1981). The
membership of DH Cep in the young open cluster NGC 7380 (Hoag and Applequist
1965; Underhill 1969) and the inferred large masses of the stellar components
make this an interesting and important binary system. Moreover, if the orbital
eccentricity found from the early radial velocity study of Pearce (1949) is
real, then DH Cep would be expected to have large apsidal motion from classical
and relativistic effects.
   Photoelectric photometry of DH Cep was conducted on 24 nights during 1985
August and September at Lines Observatory (Mayer, Arizona). The observations
were made by H. C. L. and R. D. L. using the 20 inch reflecting telescope
equipped with a 1P21 PMT and filters closely matched to the UBV system.
Differential photometry of DH Cep was made with respect to the nearby
comparison star, HD 215714 (=SAO 34785; V=+7.58; F8:) and HD 215868 (=SAO
34816; V=+8.41; B9) served as a check star. On the several nights on which
the check star was observed, no significant light variations were detected
between the check and comparison stars. Because of the angular proximity of the
comparison, check, and variable stars, the differential extinction corrections
were very small. The familiar sky-comparison-variable-comparison-sky observing
sequence was used. Differential extinction corrections were applied and the
differential measures were transformed to DeltaV and DeltaB of the standard UBV
system. The observations are plotted in Fig.1 against orbital phase using the
ephemeris,

      T_min=HJD 2446291.7784 + 2.11104dxE,

in which the period is the spectrographic orbital period given by Pearce while
the epoch is from the present observations and refers to the mid-point of the
deeper light minimum. The deeper light minimum occurs within about 0.15P of
the time of conjunction in which the more massive component is most distant
from the observer according to Pearce's spectrographic elements. It should be
recalled, however, that Pearce's study was carried out over 37 years ago and
that small errors in the published period and possible period changes could
cause the ephemeris to yield uncertain results.
   As shown in the figure, the light curves are well defined and display two
minima and two maxima within an orbital cycle. The two maxima of each light
curve are equal in brightness within +-0.005 mag while the corresponding minima
are slightly different. For the V light curve the depth of deeper minimum and
shallower minimum are 0.042 mag and 0.037 mag respectively, while for the B
observations the depths of the deeper and shallower minima are 0.045 mag and
0.039 mag, respectively.
   The small light amplitudes and the cos2Theta dependence of the brightness
variation on the orbital phase indicate that DH Cep is an ellipsoidal
variable star. As described by Morris (1985), ellipsoidal variable stars are
close binaries whose components are distorted by their mutual gravitation, but
have orbital inclinations that are too small to produce eclipses. Therefore,
the observed brightness variations occur from the changing aspects of the
tidally distorted stars as they undergo orbital motion.

                              [FIGURE 1]

   FIGURE 1. The differential B and V observations of DH Cep are
             plotted against orbital phase. The phases were computed
             with the light elements given in the text.

   A closer examination of the light curves reveals that the two minima of each
light curve are about 0.50P apart. Unless we are viewing the binary system
along its major axis so that the longitude of the periastron, omega, is nearly
equal to 90 deg or 270 deg so that ecosomega ~0.0, then it appears that the
orbital eccentricity found by Pearce is too large or even spurious.
   As shown in the figure, the shapes of the minima at 0.0P and 0.5P appear
slightly different. The deeper minimum at 0.0P is sharper and narrower than
the corresponding minimum at 0.5P in each wavelength. This behavior could be
produced by the eccentric orbit if periastron and apastron occur near the
conjunctions of the orbit at 0.0P and 0.5P, respectively. This would be
consistent with values of the longitude of periastron of omega=~90 or ~270 deg,
implied from the occurence of the two minimum nearly one half period apart.
The difference in the shapes of the two minima arise chiefly from the changes
in orbital velocity between periastron and apastron. The observed differences
in shape are small, however, and could be produced by other effects such as
circumstellar gas or streams.
   Exploratory modeling of the light curves was carried out using the Wilson-
Devinney  computer program (Wilson and Devinney 1971). We found that we could
not fit the present light curves with an orbital eccentricity greater than
about ~0.05. The preliminary modeling also indicated that both stars are
inside their Roche lobes and that the orbital inclination appears to lie
between 45 deg and 55 deg. If we adopt Pearce's values of Msin^3i obtained from
the radial velocity study, then the masses of the two components are
42M_sun<=M_1<=65M_sun and 35M_sun<=M_2<=54_sun. With a total systemic mass
of 77M_sun<=M_(1+2)<=119M, DH Cep is one of the most massive binary systems
discovered so far.
   Additional photometric observations of DH Cep are planned at Lines 
Observatory 
and also at Villanova University Observatory. These observations should
provide a check on the orbital eccentricity of the system, since, if the
orbit is eccentric, the rate of apsidal motion is expected to be quite large
from classical and relativistic effects (omega_CL+GR[theor]~30 deg/yr.).  If
this is the case, then a measurable displacement of the shallower minimum from
the deeper minimum should be observed. A more detailed analysis of DH Cep will
be made after the new light curves are obtained.


            HELEN C. LINES                 EDWARD F. GUINAN
           RICHARD D. LINES                CRAIG R. ROBINSON
          Lines Observatory            Department of Astronomy and
            P.0. Box 683                     Astrophysics
          Mayer, Arizona 86333           Villanova University
                U.S.A.                 Villanova, PA 19085, U.S.A.



References:

Hill, G., Hilditch, R.W., and Pfannenschmidt, E.L., 1976, Publ. Dom. Ap. Obs.,
   15, 1. [BIBCODE 1976PDAO...15....1H ]
Hoag, A.A., and Applequist, N.L., 1965, Ap.J.Suppl., 12, 215. [BIBCODE 1965ApJS...12..215H ]
Morris, S.L., 1985, Astrophys. J., 295, 143. [BIBCODE 1985ApJ...295..143M ]
Pearce, J.A., 1949, Astron. J., 54, 135. [BIBCODE 1949AJ.....54..135P ]
Underhill, A.B., 1969, Astron. Ap., 1, 356. [BIBCODE 1969A&A.....1..356U ]
Wilson, R.E., and Devinney, E.J., 1971, Astrophys. J., 166, 605. [BIBCODE 1971ApJ...166..605W ]
Wu, D.-C., and Eaton, J.A., 1981, Inf. Bull. Var. Stars, No. 1965.