There is an ongoing debate in the literature about the correct relationship between the free energy of transfer of a molecule i between two phases 1 and 2, DeltaGdegrees(i12), and the corresponding partition constant, K-i12(c). This work presents experimental evidence, based on data from the literature, indicating that the unit molar concentration scale is close to the correct standard state. However, a discrepancy remains that appears to be related to the free volume in the liquid phase, as has been postulated before by Vitha and Carr. This discrepancy can be quantified by an empirical linear regression with the surface tension of the respective liquid phase. Other approaches that account for an additional entropy contribution arising from differences in the molar volume of the solvents are disproved by these data. This finding has consequences for Raoult's law convention and the Flory-Huggins theory.