Definitions are TWICE as useful as standard theorems in geometric proofs. This post explains why, showing that definitions are, in a certain sense, very much like “reversible coats.”