Comment on ‘‘Translation-rotation paradox for diffusion in fragile glass-forming liquids’’

F. H. Stillinger and J. A. Hodgdon [Phys. Rev. E 50, 2064 (1994)] have proposed a ``fluidized domain'' model to explain that the diffusion coefficient ratio Dₜᵣₐₙₛ/Dᵣₒₜ increases by a factor of {~}10² on approaching the glass transition from above in fragile liquids. In their numerical estimates they make ad hoc assumptions that lead to a very small volume fraction φ₀{~}10^-5 of the fluidized domains. It is shown how these assumptions can be avoided and φ₀{~}10^-1 results from experimental values of Dₜᵣₐₙₛ, Dᵣₒₜ, and the width of the Dᵣₒₜ distribution. A modification of the Stillinger-Hodgdon model can be derived from the two-state model of Beckert and Pfeifer [Ann. Phys. (Leipzig) 16, 262 (1965)] and yields experimental estimates of a few nanometers for the fluidized domains. {} 1996 The American Physical Society.