Comment on ``Point-Contact Study of Fast and Slow Two-Level Fluctuators in Metallic Glasses'' by Keijsers, Shklyarevskii and van Kempen

We point out that a recent experiment by Keijsers, Shklyarevskii and van Kempen [Phys. Rev. Lett. 77, 3411 (1996)] on metallic glass point contacts containing two-level systems (TLS) effectively measures for the first time the conductance contributions of individual TLSs (previous experiments averaged over a distribution of TLSs). For this case the predictions of the Kozub-Kulik theory for the TLS-electron interaction differ from those of the Vladar-Zawadowski theory, the latter giving much better agreement with experiment.