5 Results and discussion

The Present data of V382 Cyg is representable only by an overcontact model with fill-out parameter of 22%. Most probably, the overcontactness of the system was formed by;

i)

orbital shrinkage due to Roche Lobe overflow from the original more massive component and

ii)

spin-orbit coupled angular momentum loss from the system, in addition to evolutionary expansion of the component stars, although radiative pressure works against the formation of contact configuration.

The results of our light curve solution are not much different from those given by Harries et al. (1997). They analyzed Landolt's photometric observations. The absolute dimensions in Table 3, which were obtained by combining our light curve solution with the radial velocity solution of Harries et al., are also not much different from those given by Harries et al. However, they are (particularly for the secondary component) 10-30 percent lower than earlier results (see e.g. Pearce 1952; Cester et al. 1978, and Popper & Hill 1991).

The mass ratio q is known to be a key parameter in the light curve solutions of eclipsing binaries. To determine the photometric mass ratio of the system, we applied a q-search procedure on our new data. Our result (q=0.68) verifies the spectroscopic mass ratios given by Popper & Hill (1991) and Harries et al. (1997). According to Leung (1988), there is very good agreement, in general, between the photometric mass ratios and the spectroscopic mass ratios derived from cross-correlation techniques. Our result confirms this conclusion.

A period study using the new data together with published times of eclipse minima revealed an increase of about 3.28 s per century in the orbital period of the system. The secular period increase of contact binary systems is possible only by mass transfer from the less massive to the more massive component. By using the conservative mass transfer hypothesis, the mass transfer rate was found to be about $5.0\, 10^{-6}\, {M_\odot/{\rm yr}}$. This rate should be the minimum value, because any mass and momentum loss from such a spin-orbit coupled contact binary system tends to decrease the orbital period. Considerable mass loss due to a strong interacting wind from V382 Cyg is inevitable. The light curve variability occuring on the rising branches of the eclipses and in the depths of the secondary minima (see Landolt 1975) and the P Cygni profiles of certain UV lines (see Koch et al. 1979) are all indicative of mass loss from the system. Koch et al. (1979) estimated a mass loss rate of $4\, 10^{-5}\, {M_\odot/{\rm yr}}$, although rates of $5\, 10^{-7}\, {M_\odot/{\rm yr}}$ are expected for O6.5 main-sequence stars (cf. Howarth & Prinja 1989).

In summary, the secular period increase of the system requires a very large rate, at least $5.0\, 10^{-6}\, {M_\odot/{\rm yr}}$, of mass transfer from the less massive to the more massive component. The mass ratio of the system is already reversed. The orbital period of the system initially decreased very rapidly and the contact configuration should have been formed before the mass ratio reversal. The period should have been increasing since mass ratio reversal of the system. The mass loss from the system should decrease the rate of period increase of the system. The short term variations in the ${\rm (O{-}C)}$ diagram should be caused by irregularities in the mass transfer and mass loss.

Acknowledgements

This work was partly supported by the Scientific and Technical Research Council of Turkey under TBAG-AY/78.