Our effective viscosity measurements³ and those of some other groups suggest that the surface mobile layer is nanometer thick and it can account for enhancements in the dynamics of films hundreds of nanometers in thickness. On the other hand, some other experiments, notably those involving gas permeability measurements and enthalpy recovery have found that confinement effect occurs on micrometer length scales. To understand the origin of these vastly different length scales, we are studying mechanical relaxations of polymer films with thicknesses from ~10 nm to ~100 microns. This thickness range will enable us to confirm if confinement effect prevails at two different length scales. If confirmed, further tests can be performed to understand the transition from one length scale to the other. Figure 3 displays representative stress-strain curves we obtained from a nanofilm. Good reproducibility of the data upon repeated runs confirms good quality of the films and measurement. ⁷