which has been hitherto completely covered, and as it were lost, again appears in the centre of the colours, asserts its right, and enables us fully to recognise the secondary nature of the accessory images which exhibit themselves as "edges" and "borders."—Note N.
247.
We can make these edges and borders as narrow as we please; nay, we can still have refraction in reserve after having done away with all appearance of colour at the boundary of the object.
Having now sufficiently investigated the exhibition of colour in this phenomenon, we repeat that we cannot admit it to be an elementary phenomenon. On the contrary, we have traced it to an antecedent and a simpler one; we have derived it, in connexion with the theory of secondary images, from