in Oct 1939, Flexner introduced the case for self-directed “theoretic” or seemingly ineffective research, citing, among the other illustrations, the abstract arithmetic of non-Euclidean geometry with no which Einstein’s principle of relativity would have been imposs