The study of V 839 Oph shows that it is an A-type W UMa eclipsing system and that the common envelope of the system has a structure with convective "continents'' on a radiative envelope. Examination of our solutions for the period variations and the light curve leads to the following conclusions.
.
variations are due to conservative
mass transfer in the system, the mass should have flown from the primary to
the secondary between 1978-1988, and from the secondary to the primary
between 1988-1998. It seems that such a frequent shift (every 
yr)
in the direction of mass transfer is not possible dynamically.
variations
is a third body in the system. For a hypothetical third body, the period
(
) and semi-amplitude (0.0065 d) of the sinusoidal
variations in Fig. 4 (click here) lead to a small mass function of f(M3)=0.00376
. We obtain 
with the assumption of circular
and co-planar orbits of the third body. The semi-major axis a3 of third
body orbit around the center of mass of the triple system is about 
.
The bolometric absolute magnitude of the third body was found to be
using the mass-luminosity function;
for 
(Demircan & Kahraman 1991).
The third body, if it exists, should revolve much beyond the outer Lagrangian
points of V 839 Oph, if it were to be stable.
is needed to produce the
observed orbital period change in the 19.62 yr cycle that is 
. The subsurface mean magnetic field of the primary then should be
about 8.21 kG. According to the discussion in photometric solution section,
if the presence of magnetic spots could create a long-term light variation
which is strongly possible but, unfortunately now, we cannot discuss the
possible relation between the long-term light curve variation and their
probable connection with the sinusoidal period variation cause of absence
enough light curves on published.
We should admit that in last few years, however the period modulation mechanism seems to be the main cause of such periodicities, the insufficient light curves on literature cannot give permit to reach final conclude in many of W UMa systems, for example, V 839 Oph (the present work), BX And (Demircan et al. 1993), AB And (Demircan et al. 1994b, Kalimeris et al. 1994), DK Cyg (Awadalla 1994) while except well observed systems, for example, SW Lac (Glownia 1986), VW Cep (Karimie 1983) and TZ Boo (Awadalla 1989).
years. At the moment we cannot
give a unique explanation for these oscillations but we have made several
suggestions. It is clear that only systematic and continuous
photometric and spectroscopic observations help to understand the connection
between light curve and the period variation.
Acknowledgements
We express our thanks to Dr. R.F. Peletier for help with the presentation. We are greatly indebted to the referees, Drs. A. Gimenez and A. Claret for valuable suggestions and helpful comments. Thanks are due to Dr. Z. Müyesseroğlu for helpful running of Wilson-Devinney code. We are grateful to the amateur astronomer, H.E. Kum, for his careful observations and participation to the observation program.