In The Aesthetics of Architecture, Roger Scruton makes at least four claims about rightness of architectural proportion. The present paper lists those claims, briefly discusses the way they are related, and, finally, selects one as the topic of discussion: the claim that there cannot be an exact, mathematical definition of rightness of proportion. Scruton's arguments for this claim are reviewed. The first is found to be substantially correct, whereas the second is found to rely on a mistaken assumption, namely the assumption that rightness of proportion is relative to a point of view. The paper ends by arguing that either the real or the apparent proportions of a building have to be definitely right, and that neither can be allowed to be definitely wrong.