THE PERIOD GAP AND MASSES OF CATACLYSMIC VARIABLES

We examine consequences of the currently-favoured view that the cataclysmic variable period gap results from a sharp decrease in orbital braking once the secondary becomes fully convective. We show that the period distribution of systems near the edge of the gap imposes, for a given stellar model, tight constraints on the parameters of the Mestel-Spruit magnetic braking law. The white dwarf mass M1 in systems such as TU Men or YZ Cnc must be high, unless these systems are assumed coincidentally to have been born close to the upper edge of the gap. With our adopted parameters, we predict M1 greater-than-or-equal-to 1 M. for both systems. Observational mass determination for these systems and those which should be discovered by ROSAT thus offer a stringent test of current theories of the period gap. Systems with 2.2 h < P < 2.8 h cannot have followed normal evolution, and must have been formed in the gap. We show that, depending on the secondary mass and age when it first comes into contact, evolution can be driven by angular momentum losses either through magnetic braking or gravitational radiation, and that for periods greater than 2.5-2.6 h, high mass transfer rates are possible, as observed in V795 Her.