The laws of the polarisation of light when reflected from the surfaces of solids and fluids, and when refracted and transmitted by translucent and transparent bodies, have been successfully investigated; but no experiments, I believe, have been made on the polarisation of light by rough or unpolished surfaces, such as ground glass, painted surfaces, pounded glass, snow, white powders, and solids and fluids reflecting white light from their interior. When studying the polarisation of the atmosphere, and anxious to discover the cause of its partial polarisation, and of the three neutral points or spots, in which there is no polarisation, I investigated the action of rough surfaces upon light, under the conviction that the sky or atmosphere was a rough surface like any aggregation of white or coloured particles. Had the atmosphere been specular like water or any body with a polished surface, the image of the sun would have been seen in it by reflexion, but being composed of aerial and aqueous molecules, it must reflect the sun's rays like pounded glass, or any white or coloured powders.