During a series of experiments in which I was lately engaged, for the purpose of determining the law of the polarisation of light, by successive reflexions from plates of parallel glass, I observed that all the images of the luminous body which were formed by more than one reflexion, were crossed by parallel fringes of coloured light, when the two plates had a small inclination to each other; and that these fringes suffered considerable changes, by varying the position of the plate with regard to the incident ray.