Polchinski has argued that the prediction of Hawking radiation must be independent of the details of unknown high-energy physics because the calculation may be performed using `nice slices', for which the adiabatic theorem may be used. If this is so, then any calculation using a manifestly covariant --- and so slice-independent --- ultraviolet regularization must reproduce the standard Hawking result. We investigate the dependence of the Hawking radiation on such a short-distance regulator by calculating it using a Pauli--Villars regularization scheme. We find that the regulator scale, $Λ$, only contributes to the Hawking flux by an amount that is exponentially small in the large variable $Λ/{T_\ssh} \gg 1$, where $T_\ssh$ is the Hawking temperature; in agreement with Polchinski's arguments. We also solve a technical puzzle concerning the relation between the short-distance singularities of the propagator and the Hawking effect.

Tex, 11 pages, no figures, new references added