A diffusive model of osmosis is presented that explains currently available experimental data. It makes predictions that distinguish it from the traditional convective flow model of osmosis, some of which have already been confirmed experimentally and others have yet to be tested. It also provides a simple kinetic explanation of Raoult's law and the colligative properties of dilute aqueous solutions. The diffusive model explains that when a water molecule jumps from low to high osmolarity at equilibrium, the free energy change is zero because the work done pressurizing the water molecule is balanced by the entropy of mixing. It also explains that equal chemical potentials are required for particle exchange equilibrium in analogy with the familiar requirement of equal temperatures at thermal equilibrium. These are topics that should be considered for inclusion in the redesign of introductory physics courses for the life sciences (IPLS). The diffusive model also makes detailed predictions for the unidirectional fluxes through single-file aquaporins that can be tested experimentally or via molecular dynamics simulation. Predictions are made for both non-equilibrium and equilibrium simulations in which there may or may not be a water chemical potential difference across the membrane. The effects of both osmolarity and hydrodynamic pressure differences are included in the model. DUE-0836833 http://circle4.com/biophysics