It is something to know what proof is and what it is not; and where can this be better learned than in a science [geometry] which has never had to take one footstep backward?

Proceedings of the American Association for the Advancement of Science, The Message of Non-Euclidean Geometry (p. 371)