Abstract

An orbital period of 0.06288(5) d has been found from a radial velocity study of the H? emission line. In addition, we have detected an extra line emitting source located ? 80° apart from the vector joining the secondary-primary centers, as measured in the opposite sense to the binary rotational motion. This is not the expected location for the hotspot in dwarf novae. This anomaly could be removed by assuming a line emission lagging behind the white dwarf binary motion. In addition, we have estimated line emissivity (? r-?) and disk radius (R = rin/rout) for 8 SU UMa stars. Most stars fit ? = 1.8 ± 0.1 but AK Cnc and WZ Sge strongly deviate from the mean; their emission line shapes can be explained assuming a post-outburst accretion disk mostly emitting close to the white dwarf (AK Cnc) and a ring-like disk (WZ Sge). In addition, we have found a tendency of long-supercycle length SU UMa stars to show very compact (large R; probably ring-like) accretion disks. If the supercycle length were basically controlled by the mass transfer rate (M), the inner disk radius would be a function of M. A white dwarf magnetic field ? 5000 G is required to fit the truncation radius with the magnetosphere radius of SU UMa stars.