Representation and Similarity: Suárez on Necessary and Sufficient Conditions of Scientific Representation