The dielectric constant and frequency dependent dielectric properties have been successfully calculated from molecular dynamics simulation trajectories. The Kirkwood factor is derived from the fluctuation of the total dipole moment. My question is related to the size effect on the Kirkwood factor, since the variation of the total dipole moment strongly depends also from the size of the system. What is the limit in the system size to obtain a converging static permittivity value, mimicking an infinitely extended system in order to obtain a macroscopic quantity. However, I found a number of publications stating to be able to define local static and frequency dependent permittivity values in anisotropic systems ? How is it possible to reduce the linear response of a macroscopic total dipole moment to single molecules or an anisotropic layer of dipoles such as water at a surface. Thanks to anyone helping me out with my question. I might be stuck in the question related to the direct comparison of microscopic and macroscopic fluctuations. -- Emanuel