In analysing the experiences of the visual arts the first essential is to avoid the word ‘see’, a term which is treacherous in its ambiguity. If we say that we see a picture we may mean either that we see the pigment-covered surface, or that we see the image on the retina cast by this surface, or that we see certain planes or volumes in what is called the ‘picture-space’. These senses are completely distinct. In the first case we are speaking of the source of the stimulus, in the second of the immediate effect of the stimulus on the retina, in the third we are referring to a complex response made up of perceivings and imaginings due to the intervention of mental structures left behind by past experience, and excited by the stimulus. The first case we may leave out of account as a matter of purely technical interest. The degree of similarity holding between the second and third, between the first effect of the stimulus and the whole visual response, will of course vary greatly in different cases. A perfectly flat, meticulously detailed depiction of conventionally conceived objects, such as is so often praised in the Academy for its ‘finish’, may be very nearly the same from its first impression on the retina to the last effort which vision can make upon it. At the other extreme a Cézanne, for example, which to the eye of a person quite unfamiliar with such a manner of painting may at first seem only a field or area of varied light, may, as the response develops, through repeated glances, become first an assemblage of blots and patches of colour, and then, as these recede and advance, tilt and spread relatively to one another and become articulated, a system of volumes. Finally, as the distances and stresses of their volumes become more definitely imagined, it becomes an organisation of the entire ‘picture-space’ into a three-dimensional whole with the characters of the solid masses which appear in it, their weights, textures, tensions and what not, very definitely, as it seems, given. With familiarity the response is of course shortened. Its final visual stage is reached much sooner, and the stages outlined above become, through this telescoping, too fleeting to be noticed. None the less the great difference between the first retinal impression and the complete _visual_ response remains. The retinal impression, the sign, that is, for the response, contains actually but a small part of the whole final product, an all-important part it is true, the seed in fact from which the whole response grows.