According to Fermat's principle the path taken between two points by a ray of light is the path that can be traversed in the least time. So if the ray is refracted and if you take two points, one from before and one from after the refraction, the path taken between them was the one that takes the shortest time. This fact can be explained by Quantum Electrodynamics, which states that any particle propagates over all possible paths and in the end all paths except extremal cancel each other out. What I am wondering though is the following: you can use various analogies for Fermat's principle, e.g. a lawnmower, moving from a sidewalk to grass: http://web.archive.org/web/20051226...5/LectureNotes/LightOptics_files/image028.jpg But does Quantum Electrodynamics also apply to a lawnmower? I am not an expert regarding QED, but a lawnmower isn't exactly a quantum object like a photon, still if you move it from a sidewalk to grass and take two points, one from before and one from after the bend, the path taken between them was the one that takes the shortest time. Why doesn't the bend happen in any other way? Is there a different explanation for it? But if there is one, isn't it strange then that different explanations lead to the same behaviour in different cases (out of the many possible paths, the one that takes the shortest time is taken - in the case of the refracted ray of light and in the case of the lawnmower)? Maybe someone can help regarding this issue as I am wondering about this for quite some time now.