RAPID OSCILLATIONS IN CATACLYSMIC VARIABLES .8. YY DRACONIS (= 3A 1148+719)

We have found a 16th magnitude cataclysmic variable, presumably a dwarf nova, near one of the positions permitted for the hard (2-10 keV) X-ray source 3A 1148 + 719. The location is also coincident with the reported position of the variable star YY Draconis, cataloged as an "Algol-type" variable. Since no Algol-type system can now be found in the vicinity, it is extremely likely that the published classification of YY Dra is incorrect and that the cataloged variable is the cataclysmic binary. A serendipitous Einstein observation of the star revealed a fairly strong hard X-ray source, making it nearly certain that this is the correct identification of 3A 1148 + 719. The optical and ultraviolet spectra of YY Dra are fairly typical of cataclysmic variables, except that the TiO absorption bands of an M dwarf can be seen for lambda > 5000 angstrom. This shows that the accretion disk is intrinsically quite faint in quiescence, suggesting a very low accretion rate. Yet the fairly blue continuum slope in the vacuum ultraviolet indicates the presence of a small hot object, presumably the white dwarf. The radial velocities at H-alpha indicate a 4.0 hr orbital period, and the infrared light curve appears to show the expected `double-humped" waveform from the distorted secondary. High-speed photometry in the U band reveals a stable periodicity of approximately 1% semiamplitude at a period of 275 s, with some power also in the subharmonic at approximately 550 s. This identifies the star as a member of the "DQ Herculis" class of cataclysmic binary, in which accretion occurs onto a rapidly rotating, magnetized white dwarf. However, the photometric period in blue light, at a time when the star was slightly fainter, was observed to be 266 s, an orbital sideband of the 275 s period. The most likely interpretation is that the shorter period is the true sidereal rotation period, while the synodic period of 275 s appears prominently in the U band due to the contribution of emission lines and the Balmer continuum.