George Berkeley argues that vision is a language of God, that the immediate objects of vision are arbitrary signs for tactile objects and that there is no necessary connection between what we see and what we touch. Thomas Reid, on the other hand, aims to establish a geometrical connection between visible and tactile figures. Consequently, although Reid and Berkeley's theories of vision share important elements, Reid explicitly rejects Berkeley's idea that visible figures are merely arbitrary signs for tangible bodies. But is he right in doing so? I show that many passages in Berkeley's work on vision suggest that he acknowledges a geometrical connection between visibles and tangibles. So the opposition between the arbitrariness Berkeley defends and a geometrical connection cannot be as universal as Reid thinks. This paper seeks to offer a plausible reading of Berkeley's theory of vision in this regard and an explanation of why Reid interprets Berkeley differently.