There are many ways in which we use unrealistic representations for modeling the physical world. In some cases we construct models that we know to be false-but not false in the sense that they involve idealization or abstraction from real properties or situations; most models do this. Instead, they are considered false because they describe a situation that cannot, no matter how many corrections are added, be physically true of the phenomenon in question. Maxwell’s ether models are a case in point. No one understood or believed that the structure of the ether consisted of idle wheels and rotating vortices, yet those types of models were the foundation of Maxwell’s fi rst derivation of the electromagnetic fi eld equations.