ABSTRACT
Scientific theories are not memory-saving devices, nor is the primary function of abstraction to economize thought. The aim of these theories is understanding; the function of abstraction is to reveal general characters. The scientist seeks form; such economy of thought as results from the discovery of form is a by-product of the scientist's activity. It is not difficult to see that Newton's theory is a constructive description. It is possible here only to enumerate some of the facts that are connected in the Newtonian system in a way that renders their connexion intelligible. In spite of its enormous interest for students of scientific method, the author cannot here follow the development of Einstein's theory of relativity. That an extremely delicate experiment should necessitate such a result is possible only at an advanced stage of a scientific theory, that is, only where such a scientific theory is a constructive description.
