Cataclysmic Variable Primary Effective Temperatures: Constraints on Binary Angular Momentum Loss

We review the most decisive currently available measurements of the surface effective temperatures, T _eff, of white dwarf (WD) primaries in cataclysmic variables (CVs) during accretion quiescence, and use these as a diagnostic for their time-averaged accretion rate, < \dot{M} >. Using time-dependent calculations of the WD envelope, we investigate the sensitivity of the quiescent T _eff to long-term variations in the accretion rate. We find that the quiescent T _eff provides one of the best available tests of predictions for the angular momentum loss and resultant mass-transfer rates which govern the evolution of CVs. While gravitational radiation is completely sufficient to explain the < \dot{M} > of strongly magnetic CVs at all P _orb, faster angular momentum loss is required to explain the temperatures of dwarf nova primaries (nonmagnetic systems). This provides evidence that a normal stellar magnetic field structure near the secondary, providing for wind launching and attachment, is essential for the enhanced braking mechanism to work, directly supporting the well-known stellar wind braking hypothesis. The contrast in < \dot{M} > is most prominent for orbital periods P _orb > 3 h, above the so-called period gap, where < \dot{M} > differs by orders of magnitude, but a modest enhancement is also present at shorter P _orb. The averaging time which < \dot{M} > reflects depends on < \dot{M} > itself, being as much as 10⁵ years for low-< \dot{M} > systems and as little as 10³ years for high-< \dot{M} > systems. We discuss in some detail the security of conclusions drawn about the CV population in light of these time scales and our necessarily incomplete sample of systems, finding that, due to the time necessary for the quiescent T _eff to adjust, the consistency of measurements between different systems places significant constraints on possible long-timescale variation in \dot{M}. Measurements for nonmagnetic systems above the period gap fall below predictions from traditional stellar wind braking prescriptions, but above more recent predictions with somewhat weaker angular momentum loss. We also discuss the apparently high T _eff's found in the VY Scl stars, showing that these most likely indicate < \dot{M} > in this subclass even larger than predicted by stellar wind braking.