The evaluation method of dielectric data based on a Williams product ansatz [A. Arbe, D. Richter, J. Colmenero, and B. Farago, Phys. Rev. E 54, 3853 (1996)] is not suited to the crossover region of the dynamic glass transition. This method is insensitive to the shape, position, or even existence of at least one of the relaxation functions if, in the extrapolation from below to above the crossover region, the two processes cross or approach one another. Since a crossover region without any particular change in the relaxation mechanism of these processes can in fact be adjusted by this method, better dielectric data for polybutadiene in the gigahertz frequency region are needed to determine whether or not there are peculiarities in the dielectric function. It would be an odd assumption that the many particularities in the crossover region in other instances are not reflected in the dielectric response.

• Received 19 December 1996

DOI:https://doi.org/10.1103/PhysRevE.60.1099