V Sge: a hot, peculiar binary system

Five sets of mean UBV light curves of V Sge covering 2 mag of its large sca le variations are analyzed. The mass ratio adopted in the analysis q = M-2/ M-1 = 3.76 is that obtained by Herbig et al. (1965) from radial velocity cu rves based on fluorescent O III lines (arising from the surfaces of the two components). Models with an accretion disk around the white dwarf primary (or a very mas sive neutron star secondary) fail completely to reproduce the shapes of the observed light curves. Successful solutions are obtained with a model involving contact configurat ion, modified by the presence of a hot, gaseous envelope (needed to explain the behavior of colors and the variable depth of the primary eclipse). Ar inclination i approximate to 71 degrees the resulting masses of the compone nts are: M-1 approximate to 0.9 M-circle dot and M2 approximate to 3.3 M-ci rcle dot. In the faintest state the secondary is a main sequence star with R-2 approximate to 1.2 R-circle dot and T-2 approximate to 12 000 K, while the main parameters of the primary are: R-1 approximate to 2.1 R-circle dot . T-1 approximate to 70 000 K, and L-1 approximate to 1 x 10(38) erg/s. Due to the high radiation pressure from the primary an expanding gaseous envel ope is formed, leading to the mass outflow from the system. Large scale variations involve significant increase of the temperatures of both components, up to about 140 000 K for the primary and about 50 000 K f or the secondary, and a considerable thickening of the gaseous envelope, wh ich contributes up to 20-30% of the total UBV flux. These variations are in terpreted as being due - in part - to the instability and large variations in the rate of mass outflow from the secondary. No obvious explanation, how ever, is offered for the major increase of the temperature and luminosity o f the primary component in the brightest state. The temperature of the primary in the faint and intermediate states (T-1 ap proximate to 70 000 K) is too low to explain the supersoft X-ray flux (obse rved only during those states), the only alternative being that it must com e from the envelope surrounding the two stellar components. Such a hypothes is can also explain the origin of the O III and O VI lines. The distance and interstellar reddening, resulting from the solution, are d = 4 kpc and EB-V approximate to 0.30-0.36 mag. The far ultraviolet fluxes, calculated with model parameters obtained from the solution, do not agree with the observed IUE fluxes, corrected for interstellar extinction using t he standard extinction law. The agreement becomes satisfactory, however, wh en arbitrarily chosen examples of non-standard extinction curves are used i nstead.